• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    In order to check the compliance of a data system during the calibration of a technical system, it is provided that the data envelope (D) is represented by a function f (zj) = and a given center (c) of the data points (xn), or f (zj) =, with a radial basis function () and coffeients (cn), and the test point (zj) is considered to be within the data envelope (D), if the condition f (zj) = <cos or f (zj ) = with predetermined opening angle () or given contour line (f *).
    公开号:AT518676A1
    申请号:T50450/2016
    申请日:2016-05-17
    公开日:2017-12-15
    发明作者:Ing Nico Didcock Dipl；Dipl Ing Dr Techn Hametner Christoph；Dr Jakubek Stefan
    申请人:Avl List Gmbh；
    IPC主号:
    专利说明:

    Method for calibrating a technical system
    The subject invention relates to a method for calibrating a technical system, which is controlled by a number of control variables and adjusts an operating point in the form of a number of state variables depending on the control variables, wherein in the calibration in an ith operating point by optimization under Compliance with predetermined constraints the control variables are sought, which are optimal in terms of at least one output of the technical system, is checked with a constraint of the optimization, whether a calculated in the optimization test point with control variables within a data envelope to a number of existing data points of the technical system.
    The calibration of an internal combustion engine is generally about setting certain predetermined control variables of the internal combustion engine in dependence on state variables of the internal combustion engine, so that certain specifications, such as, for example, are set. Emissions limits or consumption limits (general output variables), are complied with and thereby impermissible operating conditions of the internal combustion engine can be avoided. Control variables are variables by which the internal combustion engine is controlled, for example an ignition point, an injection time (eg pre-injection, post-injection), the amount of recirculated exhaust gas of an EGR (Exhaust Gas Recirculation), the position of a throttle valve, etc. State variables are variables of Internal combustion engine, depending on the control variables and depending on external influences (such as a load, environmental conditions, etc.) set. Typical state variables of an internal combustion engine are the speed and the torque. The state variables depict an operating state of the internal combustion engine. The state variables may be measured, but may also be e.g. calculated from models, from other measured variables of the internal combustion engine. The control variables and the state variables together form the input variables of the calibration, in the form of an input variable vector of all control variables and state variables. Illegal operating conditions are determined, for example, by physically measurable output variables, such as e.g. Consumption, emission (NOx, COx, soot, etc.), cylinder pressure, engine temperature, etc. are detected. As a rule, corresponding limit values of the measurable output variables are specified. The output variables are set as a function of the current operating state (state variables) and the current control variables as a reaction of the internal combustion engine. An inadmissible operating state of the internal combustion engine can also manifest itself on effects such as knocking, misfiring, etc. Such impermissible operating states occur in certain combinations of input variables. When calibrating an internal combustion engine, it is now necessary to set the control variables for given state variables so that no impermissible operating states occur and at the same time certain output variables (targets of the calibration) are optimized (usually minimized), typically emission values (NOx, COx, HC (Hydrocarbon) share in the exhaust gas, soot, particulate matter, etc.) and consumption. Calibration room is defined as the room for multidimensional (according to the number of control variables) space defined by the control variables. All permissible control variables of the test room form the drivable area within which the control variables must lie for given state variables in order not to cause any impermissible operating states. The mobile area, which depends on the current state variables, is thus a subspace of the test room. The outer edge of the mobile area is often referred to as the driveability limit.
    The control variables are often stored in control units, for example as maps depending on the state variables. For an internal combustion engine, the control variables are e.g. deposited an engine control unit.
    Today's internal combustion engines have a variety of control variables that are set in response to a variety of state variables. Due to the resulting large number of input variables, the usually non-linear influences of the control variables on the operating states of the internal combustion engine and also due to the, even multiple, dependencies of the input variables (because the driveability limit must be met) calibration is a very complex task manually can hardly be accomplished. Calibration is therefore often solved as an optimization problem, with the inputs being optimized (minimized or maximized) for one or more outputs. During optimization, the control variables are varied for given state variables in order to optimize the output variables. As output variable, emission values or consumption are frequently used in the calibration of an internal combustion engine. At the same time, additional conditions are specified during optimization, in particular compliance with the driveability limit. That is, in the optimization, the control variables may only be within the driveability limit. In this way, maps are usually created for the control variables in dependence on the state variables that are stored in the engine control unit of the internal combustion engine for controlling the internal combustion engine. In the ferry mode, the current state variables are recorded and the settings for the control variables to be made are taken from the maps.
    One problem with this is that only a limited, discrete number of data points of the input variables (one data point in this case is a concrete input variable vector) is available for the calibration. These data points are either calculated from models or measured on a test bench on the existing and operated internal combustion engine. But you have only a few points of the drivability limit available. Whether a control variable newly determined from the optimization, in most cases a control variable vector with a plurality of control variables, is within the drivability limit, or outside, can not be determined on the basis of the existing discrete data points. Therefore, so-called data envelopes have been used, which calculate an envelope of the data points, the so-called data envelope, from a limited number of data points. The driveability limit is thus formed by the data envelope. To solve this problem, various methods have been proposed.
    A first known approach is the use of so-called classifiers. A classifier assigns each data point in the test room either the "valid" or "invalid" attribute. There are several model structures for classifiers for determining the driveability limit, e.g. Support vector machines such as in G. Kampmann, et al., Support Vector Machines for Design Space Exploration, Proceedings of the World Congress on Engineering and Computer Science 2012, 24-26. October 2012, Vol II. However, this requires data points from both classes, ie both valid and invalid data points. When calibrating an internal combustion engine, however, it is usually the case that only valid data points (ie data points within the driveability limit) are available, which is why classifiers are generally unsuitable for this application.
    Therefore, shell algorithms have already become known which calculate the data envelopes exclusively from the available permissible data points. An example of this is the convex data envelope, which is defined as the smallest convex space containing the existing data points. However, the use of the convex data envelope is problematic for two reasons. First, the calculation of the convex hull is efficient only in low-dimensional experimental spaces. However, the computational effort increases exponentially with the dimension of the test room, but also with the number of data points, so that a calculation from approximately 10-dimensional test rooms is already practically impossible. Second, in practice, the relevant areas of the experimental space are often not convex, resulting in unrealistic data envelopes. This problem is often solved by more complex triangulations. A triangulation is the division of the convex data envelope into simple forms, that is, in the two-dimensional case, for example, the triangulation of the data points in triangular networks. An example of non-convex triangulations are the so-called alpha shapes, as described, for example, in H. Edelsbrunner, et al., "Three Dimensional Alpha Shapes", ACM Transactions on Graphics, Vol. 1, Jan. 1994, pages 43-72. However, the calculation of these triangulations is for higher ones
    Dimensions, and also for a high number of data points, also very complex and expensive.
    However, 10 to 20 control variables (dimension of the test room> 10) are generally to be processed, especially when calibrating an internal combustion engine. Similarly, 105-106 available data points are not uncommon for calibration. Both make the calculation of a convex data envelope during calibration virtually impossible, at least with reasonable effort and within a reasonable time.
    The above problems occur not only in the calibration of an internal combustion engine as a technical system, but of course can be transferred to any technical system, the technical system is controlled based on a specification of control variables and depending on the specification of the control variables a specific state variables , and possibly in dependence on external influences, assumes a defined state. In the calibration of the technical system, the goal is to specify the control variables as a function of the state variables, and optionally as a function of external influences, so that as impermissible operating states as possible occur and at the same time optimized output variables of the technical system with respect to an optimization target. During calibration, data fields or data tables are usually created, with which the technical system can be controlled. As a rule, these data fields or data tables are stored in a control unit of the technical system in order to be able to read out the optimum control variables for the respective operating state during operation. Typical examples of components of a vehicle that are controlled with associated controllers include, in addition to the internal combustion engine, a transmission, a traction battery, brakes, a powertrain, a suspension, etc.
    However, the calibration according to the invention can also be used to optimize the behavior of a vehicle or a component of a vehicle. Often, the driving characteristics of a vehicle (e.g., noise, suspension, damping, shifting behavior, etc.) are optimized for desired characteristics by calibration. Examples include a damping optimization, gearbox optimization, clutch optimization. For example, the stiffness of a chassis can be optimized by varying the spring parameters (control variables) in wheelbase bearings to accommodate certain output quantities, e.g. the driving dynamics or driving comfort to influence or optimize. Another example is a hydrodynamic clutch wherein the clutch fill history is optimized.
    The calibration is not limited to components of a vehicle, but can be applied in general for any machine components as a technical system.
    As a technical system for the calibration, therefore, in particular every component of a machine, e.g. a vehicle, which can be influenced by control variables and which is to be optimized with regard to a specific behavior as a function of an operating state.
    It is therefore an object of the present invention to provide a method for calibrating a technical system, which makes it possible, on the basis of valid data points, to take into account compliance with a data envelope in a simple and rapid manner.
    This object is achieved with the features of independent claim 1. The modeling of the data envelope according to the invention makes it possible to check whether an optimized test point lies inside or outside the data envelope with little computational effort. This makes it possible to carry out this check even in test rooms of high dimension (> 10), as is typically the case in the calibration of technical systems.
    When using different aperture angles cpn for different data points xn or
    radial basis functions and the interpolation condition <1> c = t with the vector t = [ti ..... tN] T with t, e {-1, 0, 1}, data points outside the data envelope can also be taken into account Modeling the area within the data envelope more accurately.
    The subject invention will be explained in more detail below with reference to Figures 1 to 3, which show by way of example, schematically and not by way of limitation advantageous embodiments of the invention. It shows
    1 shows the illustration of a conical data envelope,
    A conical local and global data envelope and
    3 shows the illustration of a data envelope with a radial basis function.
    To begin with, there are well-known methods to create a design (also known as Design of Experiments (DoE)) in the form of a set of control variables that can be used to ideally stimulate the technical system to provide static and dynamic characteristics of the technical system. If the technical system is stimulated by the control variables and the state variables are measured, one obtains data points xn, ie in this case concrete measurements on the technical system. Likewise, there are well-known methods for identifying a model of a technical system, or an output variable y of the technical system, from available data points xn. It is also possible to use known existing models for technical systems. With the model, the output quantity y no longer has to be measured, but can also be calculated for other input variable vectors from the model. Likewise, with a model, the state variables can be modeled as a function of the control variables, which makes it possible to determine further data points xn from the model. Often, the experimental design and a model are also determined simultaneously. Examples of this can be found in WO 2012/169972 A1 or WO 2014/187828 A1. For the present invention, however, it does not depend on the specific way in which one arrives at the data points xn, but it is assumed that a multiplicity of valid data points xn (ie within the driveability limit) are present. Likewise, it is irrelevant to the invention how the output variable y is determined, for example by measurement or from a suitable model.
    The invention will now be described as a technical system without restriction of generality using the example of an internal combustion engine. It is assumed that a multiplicity n = 1,..., N of valid (ie within the driveability limit) data points x "of the technical system is present. The data points xn can be determined by measurement on a test bench for the technical system, for example on an engine test bench for an internal combustion engine, or on the basis of an existing or identified model of the technical system. In both cases, it is usually the goal to cover the test room, ie the space of the control variable un, as well as possible and evenly within the driveability limit as a function of the state variables vn.
    A data point xn is in each case an input variable vector x in the form of concrete values of all control variables u and concrete values of all state variables v, that is to say xn = [u ^, v ^] T.
    The control variables un and state variables vn are also summarized in corresponding vectors.
    Calibration is then an optimization problem that is generally general in shape
    u (v) = arg min y (x) VveV
    U h (x) <0 g (x) - o can be written to. Where V denotes the state space given by the state variables v. h and g are given constraints of optimization. Optimized are output quantities y of the technical system, which are combined in an output variable vector y, for example emission values, the consumption, etc. of an internal combustion engine. It is optimized for at least one output quantity y by varying the control variable vector u in dependence on a given state variable vector v. With the constraint h, e.g. compliance with permissible operating conditions of the technical system, such as emission or consumption limits, maximum cylinder pressure, maximum engine temperature, prevention of knocking, etc., can be specified, and with the constraint g, a data envelope, e.g. the observance of a driveability limit, be specified.
    The calibration is usually in each case in a predetermined, fixed i-th, i = 1 ..... 1 operating point of the technical system, which is given by a given i-th state variable vector y, which contains all state variables v, carried out. The calibration at fixed operating points facilitates the calibration, since with it the variably i-th input variable vector Xi for the calibration to x; = [u; ] T can be simplified.
    In the same way, the constraints on hj (Uj) and gi (Ui) are simplified. In this case, we also speak of local calibration, since only the control variables u i, k, which are valid for the respective i th operating point are taken into account, but not the state variables V j. Thus, only a local data envelope is considered. The local data envelope thus only includes valid data points xni for the captured i-th operating point. The above optimization problem for the ith operating point then simplifies to Z; = argminyi (ui) u hi (u;) <0 g; (ui) - 0 where z, the result of the optimization for the state variables v, denotes. This optimization problem is solved for at least one, preferably for each, of the i operating points. Zi then designates the optimized control variables for the current ith operating point. The data points xn are only needed for the calibration of the data envelope in the form of the constraint gi (Ui). The disadvantage of this is that thereby known information of adjacent operating points is systematically ignored.
    Therefore, in the calibration, the state variables v, are often taken into account as an additional input, resulting in an input variable vector x, for calibration in an ith
    Operating point v, to x; = [u ^, v;] yields. During calibration, in turn, the control variable vector Ui is varied in dependence on a given state variable vector Vj. In the same way, the constraints at the i-th operating point v result in h (Ui, Vj) and g (Ui, Vi). In this case one also speaks of global calibration and with the constraint g (Ui, Vj) a global data envelope over all data points xn is considered. The optimization problem then follows for each of the i operating points Vj in the form zi (vi) = argminyi (ui, vi) uh (Ui, Vi) <0g (Ui.Vi) <0 where Zi (Vi) is the result of optimization for denotes the state variables Vj. This optimization problem is solved for at least one, preferably for all, i operating points. The calibration is thus also carried out at a certain i-th operating point Vj, only information from further, in particular adjacent, operating points is now taken into account via the boundary conditions h and g. As a result, the models of the output quantities y identified by the optimization become more accurate and the boundary conditions g given by the global data envelope become less restrictive, as will be explained below.
    The output quantities y, yi, respectively, of the technical system are each an optimization target or a target of the calibration, for example the minimization of the emissions, such as e.g. NOx, COx, HC or soot emission of an internal combustion engine, minimizing the consumption of an internal combustion engine, etc. The relationship between outputs y, and yj, and the control variable u and possibly the state variable v is identified by the calibration. There are well-known methods for solving the above-mentioned optimization problems, for example, such methods are described in WO 2013/087307 A2. In the optimization, therefore, the control variables u (test point z) are sought, which are optimal with regard to at least one output variable y of the technical system, the optimization aiming at a minimization or maximization. Generally, the optimization is performed iteratively until a defined termination criterion is reached. One possible termination criterion is e.g. a certain number of iterations or reaching or falling below a target of the output quantities y, y respectively. The control variables u at the time of occurrence of the abort criterion are regarded as optimal control variables u. In this case, the compliance with the secondary conditions is checked in each iteration step. For the subject invention, it does not matter how to solve the optimization problem.
    During the calibration, a new test point zj = [uj,] T or zi = uj is determined by the optimization in each iteration step j of the optimization for each of the i operating points Vj, taking into account the N existing data points xn for the constraint g. If the predetermined termination criterion is reached and the new test point Zj fulfills all secondary conditions, then the last test point Zj becomes the optimized control variable z, in the ith operating point. In order to check compliance with constraint g during calibration, it is checked in each iteration step j whether this new test point z, within the data envelope, e.g. within the driveability limit, lies, or outside. This check must also be possible sufficiently quickly for high dimensions (> 10) of the test space defined by the control variable u. For this, the data envelope D must first be defined.
    In a first approach according to the invention, a conical data envelope is modeled by a function f (zj). The new test point Zj lies outside the data envelope when the condition
    is satisfied. The parameter φ models the size or the volume of the data envelope. As a rule, the Euclidean norm | used, although other standards are conceivable.
    Here, the unit vector a is
    defined, with a center c of the data points xn.
    The center c may e.g. the center of gravity of the n = 1 ..... N data points xn according to
    may be a local center of a subset of the existing data points xn or may also be an arbitrarily chosen point. Other definitions for the center c are also possible. The vector a is thus a unit vector, which from the center c in
    Direction of the test point z, shows. In the same way
    a unit vector derived from
    Test point Zj points in the direction of an existing data point xn. The scalar product of two vectors is known to indicate the angle between the two vectors. Thus, one can imagine a cone pointing away from the center c with a peak at the test point Zj and with an opening angle a defined by the two vectors. It is thus checked by the above condition whether the largest aperture angle a = arccos (f (zi)) resulting between the data points xn and the test point z, is smaller than a predetermined pitch angle φ. If so, then the test point Zj is separated from the data points xn by a divisional cone having an opening angle α equal to the pitch angle φ, i. that no data point xn is within this divisional cone. If this is the case, then the test point z, as viewed outside the data envelope.
    This will be explained with reference to FIG. 1 on a simple example in the form of two-dimensional data points xn. The data envelope D is shown on the left for a pitch angle φ = 90 ° and on the right for a pitch angle φ = 45 °. In each case, five data points xn, n = 1,..., 5 are shown. The center c has been set as the center of gravity of these five data points xn. The data envelope D is visualized in each case in FIG. For visualization, the test data points z were determined for which the largest aperture angle α = φ. The visualization of the data envelope D is only illustrative. For a calibration, the calculation of the data envelope D is not necessary, but it is only necessary with the above condition to check whether a test point Zj lies inside or outside the data envelope D. In Fig. 1 on the left is a first test point za within the data envelope D, because there is no cone in the direction center c with opening angle φ, which separates the test point za from the data points xn (the data points X2 and X3 are within the cone). A second test point, for example, lies outside the data envelope D, because there is a cone with a tip in the direction of the center c with an opening angle φ separating the second test point zb from the data points xn (none of the data points xn lies inside the cone). FIG. 1 also shows the influence of the pitch angle φ on the data envelope D. In particular, the volume of the data envelope D is determined by the pitch angle φ. For the calculation of f (Zj) (and thus indirectly for the data envelope D), either all N existing data points xn may be used, or only a number Ae N of data points xn near the test point z may be used. In this case, for example, only the data points xn can be used for the calculation, which lie within a defined range around the test point Zj. For example, in the two-dimensional case, one might think of this area as a square, rectangle, circle, or any other geometric shape. Of course this can be generalized to any higher dimension. However, just the A, the test point Zj next (for example, according to the Euclidean distance) data points xn could simply be selected. If not all N data points xn are used, then the center c is preferably also determined only for the selected A data points xn.
    If all N existing data points xn are used for the calculation of f (zj), then the complexity for checking a test point z, of the order O (dN), where d is the dimension of the test point z, is. In the case where only Ae N data points xn are used, the complexity for checking a test point Zj is on the order of 0 (polylog N), where polylog denotes the known polylogarithm. The check of a test point Zj is therefore very quickly executable in both cases. In the event that the data points xn for the optimization also include the state variables v i, then the center c may be the same point for all operating points v i, or the center c may be suitably determined from the current operating point v, in the form c = c (v,) be dependent.
    In the case of the conical data envelope, it is also possible to take into account impossible test points z, which are located outside of the data envelope D. If information about areas outside the data envelope D, ie beyond the driveability limit, is taken into account in the modeling (calibration), the modeling of the area within the data envelope D becomes more accurate. This can be achieved by using different opening angles φη for different data points xn, n = 1... N, in the case of Ext, that is, for each data point xn, another opening angle φη. The new test point Zj is therefore outside the data envelope D, if the condition
    <coscpn Vn = l, ..., N is satisfied.
    FIG. 2 also explains the difference between a local and a global data envelope. A technical system is based on two state variables v = [vi, v2] T and four control variables u = [ui, u2, U3, uJT. The technical system used here is, for example, an internal combustion engine with the state variables Vi in the form of the rotational speed and v2 in the form of the torque of the internal combustion engine. As control variables are, for example, Ui in the form of injection timing, u2 in the form of the ignition timing, u3 in the form of the exhaust gas recirculation rate and U4 in the form of the throttle position. At a total of twelve operating points (state variables v), a number of control variables u combinations were set up on a test bench, and an output variable y, e.g. the consumption, measured to obtain the data points xn. In Fig.2 top left, the twelve operating points v, are shown. The three remaining diagrams each show combinations of two of the four control variables u,. In these diagrams, the data points xn lying in these planes are respectively represented, with the large points data points xn for one of the twelve operating points, e.g. the operating point at the respective smallest values of the two state variables Vi, v2, are shown. The small dots show data points xn of the other operating points.
    The solid lines in each case show the local data envelopes DL for the selected operating point. The dashed lines show for the selected operating point in each case the intersection of the global data envelope DG with the operating point. It can be seen directly from this that the constraints g defined by the global data envelope DG are less restrictive than those defined by the local data envelope DL.
    In a second approach according to the invention, radial basis functions Φ are used to determine whether a new test point Zj lies inside or outside a data envelope D. A radial basis function Φ is known to be a function for which φ (χ) = φ (| χ |), where as a rule the Euclidean norm |. | , or any other
    Standard, is used. At this time, all N data points x "become simultaneously according to the condition
    rated. For example, so-called Wendland functions can be used as radial basis functions, as described in Wendland H., "Piecewise polynomial, positive definite and compactly supported radial functions of minimal degree", Advances in Computational Mathematics, December 1995, Volume 4, Issue 1, pp 389- 396 described. Wendland functions are useful in this context because the resulting system of equations can be solved numerically very efficiently. However, this approach according to the invention for determining the data envelope D is independent of the choice of the radial basis functions.
    The function f (Zj) in turn models a data envelope D. A new test point Zj lies within the data envelope D if the above condition for the test point Zj is satisfied. f * e
    denotes the contour line which defines the data envelope D. Istf * small, then the data envelope D is large, and vice versa. If f * = 1, then the contour line lies on the data points xn. If f * = 0, then the contour line lies at a distance, which results from the selected radial basis functions, from the data points xn. The coefficients cn = [ci,, cN] T can be determined from the interpolation condition Oc = t, where the vector t = [1, ..., 1] T and <D; j = φ (χ;, χ] ) denotes the known N x N Gram matrix, ie
    This will be explained with reference to FIG. 3 on a simple example in the form of two-dimensional data points x ", which are represented as black dots. Several data envelopes D for different f * from the range [0, 1] are shown. The smaller f *, the larger the data envelope D becomes around the existing data points xn. The data envelopes D are again visualized only for the purpose of illustration by determining the test points Zj for which f (Zj) = f *. It is not necessary to calculate the data envelope D for the calibration. The test point za lies within the data envelope D because the condition is fulfilled for it. The test point zb is outside, because the condition is not met.
    With this method, it is also possible to take into account impossible test points Zj, ie lying outside the data envelope D. If information about areas outside the data envelope D, ie beyond the driveability limit, is taken into account in the modeling, the modeling of the area within the data envelope D becomes more accurate. This can be achieved by changing the interpolation condition to <D-c = t with a target vector t = [b, ..., tN] T and tj e {-1, 0, 1}. In this case, a non-mobile data point is specified with tj = -1, with t, = 1 a mobile data point and with tj = 0 areas can be specified in which there is no knowledge about the data points. The dots may e.g. be explicitly specified as boundary points.
    If Wendland functions are used as radial basis functions, then the complexity for checking a test point Zj is again of the order of magnitude 0 (polylog N), which in turn ensures rapid checking.
    权利要求:
    Claims (4)
    [1]
    claims
    1. A method for calibrating a technical system, which is controlled by a number of control variables (u) and depending on the control variable (u) adjusts an operating point in the form of a number of state variables (v), wherein in the calibration at an operating point by an optimization in compliance with predetermined constraints (h, g) the control variables (u) are sought, which are optimal with respect to at least one output (y) of the technical system, with a constraint (g) of the optimization is checked, if one in the optimization calculated test point (z,) with control variables (u) within a data envelope (D) around a number of existing data points (xn) of the technical system, characterized in that the data envelope (D) by one of the two functions

    and a pre-N given center (c) of the data points (xn), or f (Zj) = ^ cn * φ (| ζ ^ -xn | 2)> with a 11 = 1 radial basis function (φ) and coffeients ( cn), and that the test point (zj) is considered to be within the data envelope (D) if the condition



    f (Zj) = maxaT <cos φ or f (Zj) with a given Vxn opening angle (φ) or predetermined contour line (f *).
    [2]
    2. The method according to claim 1, characterized in that as center (c) of the center of gravity of n = 1, ..., N data points (x ") according to

    , a local center of a subset of the existing data points (x ") or an arbitrarily selected data point (xn) is used.
    [3]
    3. The method according to claim 1, characterized in that for different data points (xn) different opening angle (φη) are used.
    [4]
    4. The method according to claim 1, characterized in that the coefficients cn = [ci, ..., cN] T are determined from the interpolation condition <Dc = t, wherein the vector t = [1, ..., 1] T or t = [ti ..... tN] T with t, e {-1, 0, 1}.
    类似技术:
    公开号 | 公开日 | 专利标题
    AT518676B1|2018-02-15|Method for calibrating a technical system
    EP2999998B1|2018-08-08|Methods for ascertaining a model of an output variable of a technical system
    DE102012207124A1|2013-10-31|Method for operating an internal combustion engine, device for controlling and / or regulating an internal combustion engine, internal combustion engine and use of the device for operating an internal combustion engine
    DE102009021781A1|2010-11-25|Engine-operating method for calculating an engine-operating map for a vehicle&#39;s control device creates a map with a specified number of nodes while measuring data points to calculate a map value
    DE102016121338A1|2017-05-24|Time-discrete, rate-based, model-predictive control method for air path control of an internal combustion engine
    WO2014015974A2|2014-01-30|Improved performance of experiments
    DE102012018617B3|2014-03-27|Method for calculating motor characteristics, data processing system and computer program product
    DE102014213866A1|2016-01-21|Coordination of an internal combustion engine in development by map optimization of high-dimensional parameter spaces
    EP3374618A1|2018-09-19|System and method for calibrating a vehicle component
    DE102008005712A1|2009-07-23|Method and arrangement for controlling motors and a corresponding computer program and a corresponding computer-readable storage medium
    AT521422B1|2020-09-15|Procedure for calibrating a technical system
    DE102009001644A1|2010-09-23|Method and apparatus for correlating a cylinder charge and a maximum intake valve lift of an internal combustion engine
    DE102020111204A1|2021-10-28|Method for operating a control device for a motor vehicle and a corresponding control device
    AT522958B1|2021-04-15|Method and system for calibrating a control of a machine
    AT517251A2|2016-12-15|Method for creating maps
    DE102020111206A1|2021-10-28|Method for operating a control device for a motor vehicle and a corresponding control device
    DE102010040873A1|2012-03-22|Method for determining nitrogen oxide concentration in exhaust gas of internal combustion engine of vehicle, involves interpolating individual output parameters based on relationship between input parameter and discrete tuples
    EP3916495A1|2021-12-01|Method and arrangement for evaluating or preparing an evaluation of a sample of measurement data from a measurement of a plurality of workpieces
    DE102016225041B4|2019-08-01|Method for operating an internal combustion engine, control device for an internal combustion engine, and internal combustion engine with such a control device
    DE102020203135A1|2021-09-16|Procedure for quality assurance of an example-based system
    DE102020208151A1|2021-12-30|Method for controlling the production of a sheet metal component and method for producing a sheet metal component or several different sheet metal components
    DE102020111208A1|2021-10-28|Method for operating a control device arrangement and a corresponding control device arrangement
    DE102020003746A1|2020-08-27|CREATION OF CHARACTERISTICS FOR A CONTROL OF A COMBUSTION ENGINE
    DE102006046204B3|2008-07-17|Device for neuronal control or regulation in internal combustion engine, has piecewise parabolic function provided as activating function
    AT501214B1|2008-11-15|METHOD OF DESIGNING AND CONSTRUCTING COMPLEX TECHNICAL PRODUCTS
    同族专利:
    公开号 | 公开日
    WO2017198638A1|2017-11-23|
    AT518676B1|2018-02-15|
    EP3458699A1|2019-03-27|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    WO2004102287A1|2003-05-13|2004-11-25|Avl List Gmbh|Method for optimizing vehicles and engines used for driving such vehicles|
    DE102009054902A1|2009-12-17|2011-06-22|Robert Bosch GmbH, 70469|Method for setting functional parameters|
    AT510328A2|2011-12-12|2012-03-15|Avl List Gmbh|METHOD FOR EVALUATING THE SOLUTION OF A MULTICRITERIAL OPTIMIZATION PROBLEM|AT521927A1|2018-12-10|2020-06-15|Avl List Gmbh|Process for calibrating a technical system|
    DE102020206799A1|2020-05-29|2021-12-02|Mtu Friedrichshafen Gmbh|Method for the automated determination of consumption and / or emission values in at least one operating point of a characteristic diagram of an internal combustion engine on a test bench and a control device for controlling a test bench for an internal combustion engine|
    DE102020206798A1|2020-05-29|2021-12-02|Mtu Friedrichshafen Gmbh|Method for the automated measurement of a characteristic map of an internal combustion engine on a test bench and a control device for controlling a test bench|DE102004026583B3|2004-05-28|2005-11-24|Robert Bosch Gmbh|Method for optimizing maps|
    DE102009037254B4|2009-08-12|2011-12-08|Man Diesel & Turbo Se|Method and device for controlling the operation of an internal combustion engine|
    AT512977B1|2013-05-22|2014-12-15|Avl List Gmbh|Method for determining a model of an output of a technical system|AT521422B1|2018-06-21|2020-09-15|Avl List Gmbh|Procedure for calibrating a technical system|
    法律状态:
    优先权:
    申请号 | 申请日 | 专利标题
    ATA50450/2016A|AT518676B1|2016-05-17|2016-05-17|Method for calibrating a technical system|ATA50450/2016A| AT518676B1|2016-05-17|2016-05-17|Method for calibrating a technical system|
    PCT/EP2017/061658| WO2017198638A1|2016-05-17|2017-05-16|Method for calibrating a technical system|
    EP17723694.0A| EP3458699A1|2016-05-17|2017-05-16|Method for calibrating a technical system|
    [返回顶部]