DE19501315C1 - Control system for motor vehicle automatic transmission - Google Patents

Abstract

The automatic transmission monitors the driving conditions and the driving mode, i.e. the way the driver drives, to select the optimum gearchange points. The system has several gear ratios (G2,G3,G4) and several gearchange modes which are represented as functions each of which is made up of a linear combination of a set of basic functions with variable parameters. These are varied to provide an adaptive gearchange control. The basic functions are Spline basic functions using the parameters of vehicle speed and engine torque. Other basic functions are Gaussian. The nodal points etc. of the functions determine, e.g. the starting points for the gearchange.

Description

The invention relates to a method for adjusting the Gear stages of a motor vehicle automatic transmission depending speed of driving style and / or driving condition, in which active between several gear shift stages according to the corresponding one driving characteristics and / or driving state-dependent switching characteristics is switched. Each such switching characteristic shows the switching ten from one gear shift stage to another A suitable one for each change of gear Switching characteristic is fixed.

In order to better match the shift behavior of the transmission to the respective Adapting the user's driving style, it is already known in Depending on the driving style, several versions for each Specify switching characteristic fixed, z. B. each a characteristic curve version for calm, normal or sporty driving. The driver can then set his preferred driving style, after which the Ge drive control system the corresponding shift characteristic version selects. Alternatively, the vehicle may have a driving style label have a mood system in which driving style ind kative quantities are recorded and evaluated, which then the Ge Drive control system are supplied, after which this one Selects the characteristic curve version that corresponds to the detected driving style closest.  

For example, published patent application EP 0 531 567 A1 Control for a motor vehicle drive with an automatic Known gear, in which either several sets of Switching characteristics, one of which depends on the Value of a correction term is selected, or alternatively even a set of switching characteristics is provided, which then continuously according to a predetermined algorithm depending on the respective value of the correction term is changed. Of the Correction term for its part is based on the capture of the respective Load condition of the vehicle, which is evaluated by the Vehicle speed or the transmission output speed is determined.

From the patent US 5,089,963 is a transmission control system stem known that the gear shift stage to be set in each case by capturing size dependent on driving style and driving condition ß and adapted their evaluation in a neural network  to the current driving style and driving condition takes, choosing between two switching characteristic versions can be.

The invention is a technical problem of providing based on a method of the type mentioned, the one automatic, sensitive adaptation of the gear shift behavior the driving style and / or driving condition with a relatively low level Effort.

This problem is solved by a method with the characteristics of Claim 1 solved. Each switching characteristic is represented by a Linear combination of a set of basic functions with changes modeled functional parameters. From the registration drive indicators and / or driving state-indicative variables in driving operation and evaluation of these measured values is continuously information provided about the driving style and / or the driving condition. In Depending on this, the function parameters, e.g. B. the parameters for defining each basic function itself as well as the weighting coefficients that contribute to each Specify basic function for the respective linear combination, new the art determines that the switching characteristics and thus the Schaltver holding the automatic transmission to the determined driving style and / or the driving state can be adapted. This adjustment can done in very fine steps or even continuously.

In a preferred embodiment of the invention according to claim 2 basic spline functions are used so that the resulting linearly combined switching characteristics as spline functions ben. There is a start switch for each change of gear level characteristic curve and each with a spline function approximates, the nodes of the piecewise continuously dif referenced spline functions on the kinks of the associated Start switching characteristic curve and, if necessary, still on intermediate points are laid. Based on this start approximation everyone Shifting curve allows the adaptation of the gear shift behavior in terms of driving style or driving condition,  that the initially selected nodes and / or the weighting coefficients of the linear combination of spline basic functions Depending on the driving style and / or driving condition, within a predefined range lower and upper limit values can be varied. The bottom and upper limit values therefore represent limit curves for the Variation range of each switching characteristic. Especially when choosing of spline basis functions with polynomial degrees less than two, d. H. piecewise linear spline basis functions, with little Computing effort a quick, yet satisfactory Switching characteristic adaptation achieved.

In an alternative embodiment of the invention according to claim 3 are used as basic functions, radial basic functions, in particular whose Gaussian functions are used. In addition to the weighting coefficients for the linear combination of the various basic functions the position of the center point and the value of the band width of each radial basis function further variable Parameters, through their suitable selection, first a start switch Characteristic curve sufficiently precise for every switching stage change is modeled. Then the Mo won in each case dell characteristic curve in driving mode to the current driving mode se or the driving state by appropriate variation of the changeable Chen parameters adjusted.

A preferred embodiment of the invention is in the drawing shown and is described below. Here demonstrate:

Fig. 1 is a block diagram of the transmission switching behavior automatically adaptie to the driving style and the driving state in power Schaltstufeneinstelleinheit an automotive car matgetriebes,

FIGS. 2 to 4 are diagrams with switching characteristics at pickled second, third or fourth gear of the automatic switching between the second, third and fourth speed automatic transmission of FIG. 1,

Fig. 5 is a diagram illustrating the choice of support values for the approximation of a switching characteristic according to FIG. 4 by a spline function and

FIG. 6 is a diagram showing the spline function obtained by the approximation according to FIG. 5.

In Fig. 1, the structure of a gear shift setting unit is shown schematically, which as a function of the driving speed (v) and the throttle valve opening angle (θ) as a driving force-determining variable using shift characteristics (v _ÿ ; i ≠ j ε {2, 3, 4}) selects one of three possible gears (G2, G3, G4) between which the automatic transmission control can switch automatically. For each gear (G2, G3, G4), the shift stage setting unit has its own shift characteristic evaluation unit (G2A, G3A, G4A), which are arranged in parallel.

The shift stage adjustment unit carries out the process for the automatic shift stage adjustment, adapting to the driving style and the driving state cyclically, ie the shift stage adjustment unit determines the most favorable gear stage (G (t _i )) at a cycle time (t _i ) instead of the cycle time in the previous cycle ( t _i-1 ) selected gear (G (t _i-1 )), as shown in Fig. 1. For this purpose, the information about the throttle valve opening angle (θ (t _i )) present at this time (t _i ) is fed to the parallel switching characteristic evaluation units (G2A, G3A, G4A) by a corresponding sensor system. A speed information input of each evaluation unit (G2A, G3A, G4A) is guided to one of three connection points of a switch (S1), the information about the current speed (v (t _i )) via the switch (S1) being selectively ei ner of the evaluation units (G2A, G3A, G4A) can be fed. For this purpose, the switch (S1) can be controlled by a feedback signal via the gear stage (G (t _i-1 )) previously selected by the setting unit. In this way, in the situation shown in FIG. 1, that evaluation unit that belongs to the shift stage still present is that (G2A) for the second gear (G2) that is the signal provided by a corresponding external sensor about the current driving speed ( v (t _i )) supplied. The evaluation unit belonging to the gear step selected so far is then used to decide which gear step is the cheapest for further driving.

As a mirror image of the input side, the outputs of the three parallel evaluation units (G2A, G3A, G4A) are each connected to a connection of a second switch (S2). With this switch (S2), the output signal of the currently active evaluation unit is passed as output signal information of the switching stage setting unit via the switching stage (G (t _i )) which is most favorable taking into account driving style and driving condition. For this purpose, the switch (S2) is controlled synchronously with the input-side switch (S1) with the same control signal containing the information about the switching stage (G (t _i-1 )) present.

In Figs. 2 to 4 typical switching characteristics are shown in graphs in which the abscissa represents the vehicle speed are worn (v) and the ordinate represents the throttle valve opening angle (θ) on how they are used in the Schaltkennlinienauswerteeinheiten (G2A, G3A, G4A) become. The throttle valve opening angle (θ) is given as a percentage of the maximum throttle valve opening angle. Is shown in FIG. 2 in this connection that for the evaluation unit (G2A) with an engaged second gear (G2) operative Dia program, while Fig. 3 that for the evaluation unit (G3A) with gear third gear (G3) and Fig. 4 dasjeni Play back for evaluation unit (G4A) with fourth gear (G4) engaged. Each of these diagrams contains two switching characteristics (f₂₃, f₂₄; f₃₂, f₃₄; f₄₂, f₄₃) through which the Fahrge speed (v) throttle valve opening angle (θ) level in three areas corresponding to the three possible switching levels (G2, G3, G4 ) is divided. The identities f₂₄ ≡ f₃₄, and f₄₂ ≡ f₃₂ apply, since the choice of the fourth gear (G4) as the future gear stage is independent of whether the second (G2) or the third gear (G3) is present, and analogously the downshift in the second gear (G2) is independent of whether the third (G3) or the fourth gear (G4) is present. The switching characteristics (f _ÿ ; ≠ j {2, 3, 4}) shown can be evaluated by the evaluation units (G2A, G3A, G4A) during driving operation in their course in the driving speed (v) throttle valve opening angle (θ) diagram current driving style and driving condition are adapted, which will be discussed in more detail below. Figs. 2 to 4 show a state example in which the shift characteristic curves (f₂₃ to f₄₃) just have a process adapted to a more sporty driving history.

As soon as the currently active shift characteristic evaluation unit has determined the two adapted shift characteristics upon input of the current driving speed (v (t _i )), as will be described in more detail below, it determines in which of the three shift stages separated by the two current shift characteristics (G2 , G3, G4) the current measuring point with the coordinates (v (t _i ), θ (t _i )) is in the vehicle speed throttle valve opening angle diagram and selects the associated gear shift stage as a new gear (G (t _i ) ) out. This is illustrated in Fig. 2 for the case that the driving speed throttle opening angle measuring point (v (t _i ), θ (t _i )) is in the area for the third gear (G3). The evaluation unit (G2A) responsible for the second gear (G2) to date recognizes the position of the current measuring point (v (t _i ), θ (t _i )) by the fact that the measured throttle valve opening angle (θ (t _i )) between the for the current driving speed (v (t _i )) from the current switching characteristic curve (f₂₄) for switching to fourth gear generated function value (θ _iu ) and from the current switching characteristic (f₂₃) for switching to third gear for this driving speed (v (t _i )) generated function value (θ _io ). In this case, the evaluation unit (G2A) emits a control signal for switching to third gear (G3). In contrast, if the measured throttle valve opening angle (θ (t _i)) about the current supplied by the shift characteristic curve (f₂₃) for switching to the third speed at the given traveling speed (v (t _i)) upper value (θ _io = (f₂₃ (v ( t _i ))) is a signal to maintain the second gear (G2) is emitted, and if the measured throttle valve opening angle (θ (t _i )) below that by the function value of the switching characteristic (f₂₄) for a switch to the fourth Gear at the given driving speed (v (t _i )) delivered value (θ _iu = (f₂₄ (v (t _i ))), the evaluation unit (G2A) gives a signal to switch from second (G2) to fourth gear ( G4).

Subsequently, 5 and 6 with reference to FIG. The procedure for the adaptation of the shift characteristic curves (f₂₃ to f₄₃) to the respective mo mentane driving and explained the driving state exemplary of the shift characteristic curve (f₄₃) that determines as to when th from four gear to a lower Gear down. As a prerequisite for the later adaptation process, a starting form of this switching characteristic (f₄₃) is initially specified. As a start-switching characteristic, the course of the switching characteristic (f₄₃) shown in FIG. 4 is assumed without restriction of the generality in a sporty driving style. This starting characteristic is then approximated by a spline function, ie by a linear combination of a set of basic spline functions with variable and therefore adaptable parameters. For details of this conventional approximation of a function by a linear combination of a set of spline basic functions, reference can be made to the literature. As a freely selectable parameter, this type of approximation includes, in addition to the weighting coefficients, which indicate the proportion of the respective basic function, a number of archetype nodes as base values that subdivide the entire archetype area into intervals in which each spline basic function is separated by one another different polynomial function sections is given.

In application to the specific case, each node of the start switching characteristic (f₄₃) is assigned to such a node (k _m ), as is marked by the crosses in Fig. 5. In addition, further nodes can be specified between inflection points, z. B. in nonlinear functional sections, such as the further node shown in Fig. 5 (k _n ). The switching characteristic (f₄₃), of which a series of associated data points is shown as individual points in FIG. 5, can be approximated with a desired accuracy by a suitable number of spline basic functions and a suitable choice of the variable parameters. In the present case, a particularly simple, fast and efficient approximation algorithm is chosen in such a way that only spline basic functions with a polynomial degree less than two, ie piecewise linear basic functions, are selected. As a result, the spline function (S₄₃) shown in Fig. 6 is a continuous, piecewise linear approximation function for the switching characteristic (f₄₃), the linear sections of which each have two adjacent nodes (k _m , k _n ), as they are marked in Fig. 5, connects straight to each other. By selecting a sufficient number of additional nodes in non-linear function sections, these non-linear sections can also be approximated with sufficient accuracy by piecewise linear spline function sections.

It results for this special spline approximation that the weighting coefficients for forming the spline function (S ₄₃ ) from the individual spline basic functions correspond precisely to the throttle valve opening values of the selected nodes, which makes the algorithm particularly simple and efficient. Because with the choice of the nodes (k _m , k _n ), both the node original image values as the corresponding vehicle speed values and the weighting coefficients as the associated throttle valve opening angle values are fixed without these parameters having to be calculated in an additional step. In the example of FIGS. 5 and 6, twelve nodes are specified, which is why ten spline basis functions with a polynomial degree less than two are required for an exact approximation of these twelve data points. The original image values, ie the vehicle speed values, which form interval limits for the spline basic functions, result in accordance with FIG. 5 at 39.67; 39.67; 43.87; 47.55; 60.60; 80.90; 136.40; 136.40; 180.72; 180.72; 250.0; and 250.0. The associated weighting coefficient vector, which specifies the proportions of the ten spline basic functions for the spline approximation function (S₄₃) that correspond to the throttle valve opening angle values of the nodes, is in this case (6.46; 8.94; 12.0; 29.99; 36.5; 61.05; 88.0; 88.7; 91.11; and 91.11).

That way as an adaptation starting point in the form of a spline function (S₄₃) given switching characteristic (f₄₃) can be in closing driving operation with little effort to the respective Driving style and the driving condition in its course continuously suitable adjust net. For this, first of all conventional system units, which are not discussed in detail here are needed, a recording of driving style and driving status indi native sizes made, e.g. B. in addition to driving speed and Throttle valve opening angle the longitudinal and / or lateral acceleration tion, the steering angle acceleration etc. by evaluating the information obtained from these quantities then becomes information generated about the driving style and driving condition, especially the Driving dynamics characterizing driving style, the between is zero and one, with zero being the quietest and the Value one represents the sportiest driving style, and one that Road gradient factor that characterizes the driving state if between the values zero and one, where the value zero a level lane and the value one a very represents steep slope. Furthermore, the System part assessing driving style and driving condition means for Detection of driving situations in which the driver manually in the vehicle control intervenes.

The three above-mentioned information on driving style and driving status are supplied in a manner not shown to the three parallel switching characteristic evaluation units (G2A, G4A), which then adapt their switching characteristics approximated by spline functions by changing the position of the nodes (k _m , k _n ) is changed as a function of the driving dynamics factor, the gradient factor and / or the presence of a manual control intervention. In particular, it is preferably provided to change the horizontal position of the nodes (k _m , k _n ) within a respective lower and upper limit value depending on the driving dynamics factor or the gradient factor. The node limit values define limit curves for the respective shift characteristic, which the latter should not cross in order to maintain an appropriate gear shift behavior. With li nearer adaptation z. B. the node starting from the lower limit at zero for the dynamic factor or the gradient factor linearly increased with the respective dynamic or Ge gradient factor towards the upper limit, which is then reached at value one for the dynamic or gradient factor. This measure adapts the respective switching characteristic by shifting the horizontal switching characteristic. A shift of the same in the diagram to the right, i.e. in the direction of higher vehicle speed values, with a higher driving dynamics or gradient factor means on the one hand a more dynamic shift behavior, since upshifts only take place at a higher speed, and on the other hand a lower gear becomes longer for a given throttle valve opening angle keep as desired when driving down a slope.

In addition to the adaptation of the switching characteristics described above horizontal displacement makes vertical displacement possible provided the nodes and thus the switching characteristics, to the gear shift behavior in driving situations with manual Adjust the driver's intervention in the control as best as possible can. Such a vertical shift characteristic shift he follows by changing the values for the weighting factors, which, as stated above, the throttle valve opening angle values of the Correspond to nodes. Analogous to the variant described above ren of the nodes in the horizontal direction can also the ver tical change of the node location within each given limit values.

With the method described above, an automatic, stepless online adaptation of the gear shift behavior to the Driving style and driving condition through appropriate adjustment of the Gearshift characteristics setting defining shift characteristics  reached. It is understood that modifications of the above be written procedure to the extent specified by the claims put invention are possible. In particular, the method can also be used for a motor vehicle automatic transmission, which scarfed between more or less than three switching stages ten, where then the number of switching characteristics for each the shift characteristic evaluation unit assigned to the gear stage is modified accordingly. In addition, the switching shown characteristic curve evaluation units in a single evaluation unit be integrated.

The approximation of the switching characteristic curves on which the switching characteristic curve adaptation is based by a linear combination of a set of basic functions is also not limited to the spline function approximation explained, but other common algorithms for interpolation between predetermined base values are of course also considered. In particular, a basic function set consisting of individual Gaussian functions (f _i ) of the form f _i (x) = exp [- (xc _i ) ² / σ _i ²)] with a selectable center point (c _i ) and a selectable half-value width (σ _i ) can be used, whereby then, in contrast to the above spline approximation, the throttle valve opening angle as the archetype size and the driving speed as the functionally dependent size is chosen, since this enables a better gaussian function approximation to be achieved with the given shape of the switching characteristics. The center points can be divided evenly over functional sections with a relatively smooth course and can be selected to be clustered around step areas. For efficient approximation, the weighting coefficients can be set equal to the driving speed values of the selected center points, and the bandwidths can be calculated in smooth functional sections from a heuristic view of values up to, for example, the second nearest neighboring point and in the vicinity of step discontinuities e.g. B. can be set to 0.1. The variation of the switching characteristics modeled by such a Gaussian approximation can take place analogously to the procedure described above, e.g. B. can be provided to set the Fahrge speed value of each center point depending on the driving dynamics factor lying between zero and one or the gradient factor between a lower and an upper speed limit value. Similarly, the other variable parameters of this Gaussian approximation can be set appropriately.

Claims (3)

1. Method for driving style and / or driving state-dependent Ge gear shift stage setting for a motor vehicle automatic transmission, in which

• - Is automatically switched between several gear shift stages (G2, G3, G4) according to the associated driving style and / or driving condition-dependent switching characteristics (f _ÿ ; i ≠ j {2, 3, 4}), characterized in that
• - Each switching characteristic (f _ÿ ) is determined by a linear combination of a set of basic functions with variable function parameters and
• - In driving operation, driving style and / or driving state indicative quantities are recorded and the variable function parameters are adaptively redefined depending on the determined driving style and / or the determined driving state.

2. The method according to claim 1, further characterized in that spline basic functions with the driving speed (v) as the master image size and a driving force-determining variable (θ) as the image size are used as basic functions, with nodes (k _m , k _n ) thereof at least Kinks of a given start switching characteristic curve are set and the nodes and / or the weighting coefficients of the respective spline function can be changed depending on the determined driving style and / or the determined driving state within predetermined lower (θ _iu ) and upper limit values (θ _io ) . 3. The method of claim 1, further characterized in that as basic functions radial basic functions, in particular Gauss functions, with a driving force determining quantity (θ) as the original Use image size and driving speed (v) as the image size be det, the center point and / or the bandwidth depending the radial basis function and / or the weighting coefficients the respective linear combination of the basic functions in Ab dependence on the determined driving style and / or the determined driving state within the specified lower and upper limits values are changeable.