Computer-implemented method for controlling components in a motor vehicle

Abstract

Computer-implemented method for controlling components (108) in a motor vehicle, comprising the following steps: -determination of motor vehicle data (100) and environmental data (101); -Definition of a state function (102) for each of the components (108); -Determining a component state (104; 203) for the components (108) using the respective state function (102), the motor vehicle data (100) and the environmental data (101); -Optimization of the component states (106; 213), wherein the optimization of the component states (106; 213) comprises a calculation of a sum (105) of the component states and an optimization of the sum, wherein a target function (109) is defined for achieving a goal (107), wherein a target state (110) is determined, wherein the target state (110) indicates how well the goal is achieved, and wherein the target state (110) is included in the calculation of the sum (105), characterized in that the goal (107) is the addition of a further function of the motor vehicle.

Description

The present invention relates to a computer-implemented method for controlling components in a motor vehicle according to claim 1.Functions and components become more and more complex in motor vehicles. This applies in particular to the context of various functions and components. In the prior art, when a new function or a new component is introduced into a motor vehicle, relationships between this new function or the new component and the already existing functions or components are taken into account manually. However, due to an inalquacy and / or for cost and cost reasons, frequently some relationships are not taken into account in this case.WO 2004 / 066166 A2 discloses a method for simulating complex dynamic physical systems including environmental factors by a plurality of modules which supplement one another and an autonomous agent.DE 10 2018 211 575 A1 discloses a method for operating an electrical on-board power supply of a motor vehicle, wherein the on-board power supply comprises as components at least one rechargeable electrical battery and an electric motor usable as a drive motor and generator, and at least one control device assigned to at least one component is used for controlling the components according to operating parameters describing the energy consumption and energy output of the components, wherein the operating parameters are determined as output data of an artificial intelligence-using and / or artificial intelligence-programmed determination algorithm, which is determined from input data describing the current operating state of the motor vehicle, the output data.DE 10 2015 221 968 A1 discloses a method for creating an operating strategy for component systems of variable topology, wherein components of a predefined topology are coupled to one another and are represented by function blocks. In contrast, the object of the present invention is to create a method for integrating new functions or components into a motor vehicle with less manual effort. It is another object of the present invention to provide a motor vehicle into which new functions or components can be integrated with less manual effort. Furthermore, a system with a plurality of such motor vehicles is to be provided.This object is achieved by a method according to claim 1, a motor vehicle according to claim 7 and a system according to claim 8. Embodiments of the invention are set forth in the dependent claims.According to the method according to claim 1, motor vehicle and environmental data are first determined. The motor vehicle data can relate, for example, to a current shipment of the motor vehicle and can be determined, for example, using sensors. The motor vehicle data can relate, for example, to the speed of the motor vehicle, the distance of the motor vehicle from obstacles, the currently engaged gear, the clutch state, the engine temperature, the accelerator pedal position and / or a braking torque. The environmental data can relate to an environment of the motor vehicle and can likewise be determined, for example, using sensors. The environmental data can relate, for example, to an air temperature, an air pressure, an air humidity, a condition of a road and / or a floor covering.A respective state function is defined for one of the components. Each component can thus be assigned a state function. In the context of this description, a component is understood to mean in particular both an individual component of the motor vehicle, such as, for example, the engine, the battery, a pump or the cooling water, and complex components which consist of a plurality of elements. Complex components are, for example, the cooling circuit, the driver comfort, the driving behavior on a specific road surface, such as, for example, snow, or the start-stop function.Using the respective state function, the motor vehicle data and the environmental data, a component state for the components is determined. The component state may be, for example, a numerical value. The component states are optimized.In this way, for example, a new function or a new target can be integrated into the motor vehicle in a particularly simple manner. The new function or destination is defined as a component with an associated state function. The method now carries out an optimization of the component states, so that an optimization is also carried out for the new function. In this case, it is not necessary to define beforehand by a person skilled in the art in which context the new function or the new target is associated with the already present components. This is determined by the method itself, in that the component states are optimized again and again. In this way, after a sufficiently long time, the relationships between the various components become clear. For example, if parameter A is increased, this may have a positive effect on components X and Y, but a negative effect on component Z. On the other hand, if parameter B decreases simultaneously, for example, the positive effect on components X and Y may remain, but the negative effect on component Z may be decreased. This applies even to parameters in which a person skilled in the art would initially have assumed no correlation at first glance.The optimization of the component states comprises a calculation of a sum of the component states and an optimization of the sum. It can thus be achieved that not a single component state is optimized in the best possible manner, but that the entire state of the motor vehicle is always optimized.A goal function for achieving a goal is defined. This definition can be effected, for example, by a user. A target state is determined. The target state indicates how well the target is reached. The target state may be, for example, a numerical value. The target state is included in the calculation of the sum. The target function can be a function to be newly integrated into the motor vehicle, such as, for example, an extension of the service life of the battery, an increase of the maximum achievable speed, an improvement of the acceleration of the motor vehicle or an improvement of the driving behavior of the motor vehicle, in particular on a specific floor covering.Because the target state is included in the calculation of the sum, the target function is also optimized during the optimization of the component states, so that after some optimizations have been carried out, the target is reached without a person skilled in the art having to manually input or take into account a relationship between the new target function and already present components or specific parameters. In this case, the method automatically inserts the target function into the already existing system of components and after some time reaches an optimized state in which the new target has been reached to a certain degree. In this case, care is taken that after the optimization, the components already present beforehand do not have a component state that is too poor.According to an embodiment of the invention, the state function may be autonomously learned using the motor vehicle and environmental data. In the context of this description, the term "autonomous" is understood in particular to mean that the corresponding method step is carried out without manual intervention by a user. The state function is defined in such a way that it is suitable for determining the component state. If, for example, a component is adversely affected by an excessively high air temperature, the state function is learned in such a way that this behavior is mapped by it. According to one specific embodiment of the present invention, the optimization of the sum may include a change in operating parameters of the motor vehicle. A parameter can relate to a plurality of components and possibly also to the target function.According to one embodiment of the invention, the optimization of the sum can be carried out using an evolutionary algorithm. In the context of this description, an evolutionary algorithm is understood to mean, in particular, an optimization method whose mode of operation is inspired by the evolution of natural living beings. The component states and possibly also the target state represent the living beings competing for a habitat. Thus, for example, optimization of a single component state may negatively affect a different component state. Since the sum of the component states is optimized, after many optimization steps for the entire motor vehicle, a state is achieved in which each component state has combated its habitat and works sufficiently well.According to one embodiment of the invention, a threshold value can be established for each component state. When optimizing the sum, it is prevented that the component states fall below the respective corresponding threshold value. This is advantageous in that none of the component states becomes too poor during the optimization, so that the component no longer functions properly.According to one embodiment of the invention, a central digital data memory arranged remotely from the motor vehicle can be accessed during the optimization of the component states. The data memory contains information about component states of other motor vehicles. The central digital data store can be referred to as cloud data store, for example. The information stored in the central digital data memory can be used, for example, to carry out changes that have already been carried out in other motor vehicles and have led to positive results.For example, the component state of the battery could be in need of optimization, since it has deteriorated. First, it is unknown how it could be improved. For example, the central digital data store could store information indicating that certain parameters may be changed at low air temperatures to improve component state. This information may have been stored in the central digital data memory, for example, by motor vehicles, which are operated more frequently at low air temperatures due to geographical conditions.The information stored in the central digital data memory can also be used to detect redundancies. For example, if a component receives multiple input signals from multiple sensors and performs certain functions based on these input signals, problems may arise in the prior art if one of the input signals is omitted, for example because of a defect in the corresponding sensor. However, the information stored in the central digital data memory can be used to perform the function even though one of the input signals has been omitted. This may be possible, for example, since a correlation between the input signals has been established over a longer period of time and with attention to a plurality of motor vehicles, so that the function can possibly no longer be carried out in the same quality as before, but nevertheless satisfactorily.It may also be possible to realize a function in the motor vehicle that does not appear possible at all with the components of the motor vehicle according to the prior art. As in the case of the omission of an input signal, for example, the corresponding sensor cannot be present from the outset. With the above-described procedure, the function can nevertheless also be realized without this sensor.The motor vehicle according to claim 7 comprises a plurality of components and a control system for controlling the components and their functions. The control system is configured to execute a method according to an embodiment of the invention.The system according to claim 8 comprises a plurality of motor vehicles according to an embodiment of the invention. The control systems are designed to store information about the optimization of the component states in the central digital data memory. This information can comprise, for example, parameter changes and associated improvements or degradations occurring in the component states.Further features and advantages of the present invention will become apparent from the following description of preferred embodiments with reference to the accompanying drawings. This shows FIG. 1 is a schematic flow diagram of a method according to an embodiment of the invention; and FIG. 2 shows a schematic flow diagram of a method according to one embodiment of the invention.Motor vehicle data 100 and environmental data 101 are ascertained. This can be done, for example, by sensors of the motor vehicle. There is also a component 108 and a target 107. The component 108 is assigned a state function 102 that has been autonomously learned from the motor vehicle data 100 and the environmental data 101. A target function 109 is assigned to the target 107. The objective 107 is to add another function of the motor vehicle.In step 103, a component state 104 is determined from the state function 102. This component state 104 is an indicator of how well the corresponding component 108 works. It should be appreciated that component 108 may be a particular component such as the battery or motor. However, the component 108 may also be somewhat more complex or a function, such as a coolant circuit or a start-stop function. The component state 104 may be, for example, a numerical value.A target state 110 is determined from the target function 109 in step 103. The target state 110 is an indicator of how well the target 107 has been reached. The target state 110 may be, for example, a numerical value.In step 105, the sum of the target state 110 and the component state 104 is calculated. In step 106, this sum is then optimized. In this case, motor vehicle data 100 are changed, since specific parameters are changed in order to optimize the summ. When the method has been performed a number of times, the system reaches an optimized state in which the target 107 has been reached and the component 108 continues to function sufficiently well without a user having to take into account a relationship between the target 107 and the component. The target 107 has thus been reached by the motor vehicle without requiring knowledge about the relationship of the target 107 to the already existing component 108.In the method in FIG. 2, in step 201 an ecosystem is defined which comprises two component states 203. As already mentioned, the component states 203 can also be the states of specific functions of the motor vehicle. In step 204, extinction protection 204 is performed. Here, it is checked whether one of the component states 203 is smaller than an associated threshold value. If this should be the case, the corresponding component state 203 is improved as quickly as possible, since it is to be assumed that the associated component is not functioning properly.In step 205, correlation coefficients between the component states 203 are determined via a correlation analysis. From this, in step 206, common relative coefficients are determined. In step 207, an ecosystem performance value is calculated by the relative correlation coefficients. This reflects all component states 203 weighted according to their dependence.In step 208, a prediction is made about different time factors as to how the biozenosis factor develops as a function of the current actual values. Information from a central digital data memory 209 is also used in this case. This information may have been collected from other motor vehicles. For example, the information may include an indication of how the component state 203 of a particular component may be improved under certain conditions.In step 210, how much a time delay is is determined via an autocorrelation between the component state 203 and the Biozenose factor. This can be used to weight the time factors. In step 211, a ready-to-adjust state for changing the ecosystem is determined. For example, if a high Biozenosis factor prevails, a new function can be added more quickly and better than if the component states 203 are rather poor.In step 212, the sum of the component states 203 is calculated. This is optimized in step 213. The optimization results in the change of some operating parameters of the motor vehicle. When operating parameters are changed, real feedback is used as reinforcement learning to improve the method.If the changed operating parameters result in the best component state 203 being not greater than a state threshold value, these changed operating parameters are not used in the motor vehicle.If an input signal fails to fulfil a function, the method of FIG. 2 can ensure that this function is still available. This can be achieved by testing during regular operation, i.e. without failure of an input signal, whether specific functions can also be carried out without specific input signals. If an input signal then really fails, this experience may be used.The central digital data memory 209 is particularly advantageous in order to make available many data from many motor vehicles as many as possible to motor vehicles. In this way, it is also possible to determine unlikely but effective methods for improving specific component states 203.

Claims

Computer-implemented method for controlling components (108) in a motor vehicle, comprising the following steps: - determination of motor vehicle data (100) and environmental data (101); - definition of a respective state function (102) for a respective one of the components (108); - determination of a component state (104; 203) for the components (108) using the respective state function (102), the motor vehicle data (100) and the environmental data (101); - optimization of the component states (106; 213), wherein the optimization of the component states (106; 213); 213) comprises a calculation of a sum (105) of the component states and an optimization of the sum, wherein a target function (109) for reaching a target (107) is defined, wherein a target state (110) is determined, wherein the target state (110) indicates how well the target is reached, and wherein the target state (110) is included in the calculation of the sum (105), characterized in that the target (107) is the addition of a further function of the motor vehicle.Method according to one of the preceding claims, characterized in that the state functions (102) are autonomously learned using the motor vehicle data (100) and the environmental data (101).Method according to one of the preceding claims, characterized in that the optimization of the sum (105) comprises a change in operating parameters of the motor vehicle.Method according to one of the preceding claims, characterized in that the optimization of the sum (105) is carried out using an evolutionary algorithm.Method according to one of the preceding claims, characterized in that a threshold value is defined for each component state (104; 203), wherein the component states (104; 203) are prevented from dropping below the respective corresponding threshold value during the optimization of the sum (105).Method according to one of the preceding claims, characterized in that, when optimizing the component states (106; 213), a central digital data memory (209) arranged remotely from the motor vehicle is accessed, wherein the data memory (209) comprises information on component states of other motor vehicles.Motor vehicle comprising a plurality of components and a control system for controlling the components and their functions, wherein the control system is designed to carry out a method according to one of the preceding claims.A system comprising a plurality of motor vehicles according to the preceding claim, wherein the control systems are configured to store information about the optimization of the component states in the central digital data memory (209).

Images