Motor power scheduling system combined with dynamic feedback

Abstract

The invention discloses a motor power scheduling system combined with dynamic feedback, which relates to the technical field of motor control, and comprises the following steps: predefining a driving state refreshing window and a driving road condition refreshing window; obtaining real-time driving state information; performing power distribution analysis according to the real-time driving state information, and outputting distributed driving power; activating a driving road condition refreshing window according to the generation node of the distributed driving power; with the driving road condition refreshing window as a constraint, the driving road condition of the electro-tricycle is collected, and real-time driving road condition information is obtained; performing power distribution optimization on the distributed driving power according to the real-time driving road condition information to obtain optimized power distribution; and optimizing power distribution to smoothly replace distributed driving power to carry out multi-motor coordinated driving of the electro-tricycle. The technical problem of poor driving stability caused by inaccurate motor power distribution in the prior art is solved, and the technical effects of improving the motor power distribution precision and improving the driving stability are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of motor control, and particularly to a motor power scheduling system combined with dynamic feedback. Background Art

[0002] In existing electric tricycles and other electric vehicles with multi-motor drive systems, there is a problem of inaccurate motor power distribution. Generally, when an electric tricycle is in complex road conditions, such as going uphill, turning, accelerating or decelerating, the power demand of the motor will change drastically. However, the motor power scheduling systems in the prior art often cannot adapt to these changes in real time, resulting in inaccurate motor power distribution. This inaccurate power distribution may cause unstable driving performance, such as large fluctuations in vehicle speed, insufficient acceleration or unbalanced driving, etc., thus affecting driving stability. Especially in complex road conditions or sudden changes, it is easy to cause the electric tricycle to roll over. Summary of the Invention

[0003] This application provides a motor power scheduling system combined with dynamic feedback, which is used to solve the technical problem of poor driving stability caused by inaccurate motor power distribution in the prior art.

[0004] In view of the above problems, this application provides a motor power scheduling system combined with dynamic feedback.

[0005] This application provides a motor power scheduling system combined with dynamic feedback, and the system includes: A predefined module, which is used to predefine a driving state refresh window and a driving road condition refresh window; a driving state recognition module, which is used to perform the recognition of the driving state of the electric tricycle with the driving state refresh window as a constraint to obtain real-time driving state information; a power distribution analysis module, which is used to perform power distribution analysis according to the real-time driving state information and output distributed driving power; a driving road condition refresh window activation module, which is used to activate the driving road condition refresh window according to the generation node of the distributed driving power when using the distributed driving power for the multi-motor coordinated driving of the electric tricycle; a driving road condition acquisition module, which is used to perform the acquisition of the driving road condition of the electric tricycle with the driving road condition refresh window as a constraint to obtain real-time driving road condition information; a power distribution optimization module, which is used to optimize the power distribution of the distributed driving power according to the real-time driving road condition information to obtain optimized power distribution; a multi-motor coordinated driving module, which is used to perform the multi-motor coordinated driving of the electric tricycle by smoothly replacing the distributed driving power with the optimized power distribution.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages: This application predefines a driving state refresh window and a driving condition refresh window; constrained by the driving state refresh window, it identifies the driving state of the electric tricycle to obtain real-time driving state information; performs power distribution analysis based on the real-time driving state information and outputs distributed driving power; when using the distributed driving power for the multi-motor coordinated driving of the electric tricycle, activates the driving condition refresh window according to the generation node of the distributed driving power; constrained by the driving condition refresh window, collects the driving conditions of the electric tricycle to obtain real-time driving condition information; optimizes the power distribution of the distributed driving power according to the real-time driving condition information to obtain optimized power distribution; uses the optimized power distribution to smoothly replace the distributed driving power for the multi-motor coordinated driving of the electric tricycle. The present invention solves the technical problem in the prior art that the motor power distribution is inaccurate, resulting in poor driving stability. By predefining the driving state refresh window and the driving condition refresh window, combining real-time driving state and road condition information, and adopting a multi-stage power scheduling and optimization method, it realizes the multi-motor coordinated driving of the electric tricycle. After the driving state is identified, power distribution analysis and optimization are carried out, and the distributed driving power is replaced through smooth transition calculation to ensure smooth driving performance under different road conditions, achieving the technical effects of improving the accuracy of motor power distribution and enhancing driving stability. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0008] Figure 1 It is a schematic structural diagram of a motor power scheduling system with dynamic feedback provided for an embodiment of this application.

[0009] Figure 2 It is a schematic flow chart of the execution steps of the power distribution analysis module in the motor power scheduling system with dynamic feedback provided for an embodiment of this application.

[0010] Description of reference numerals: Predefined module 11, driving state identification module 12, power distribution analysis module 13, driving condition refresh window activation module 14, driving condition collection module 15, power distribution optimization module 16, multi-motor coordinated driving module 17. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0011] This application provides a motor power scheduling system combined with dynamic feedback, aiming to solve the technical problem in the prior art that inaccurate motor power distribution leads to poor driving stability. By predefining a driving state refresh window and a driving road condition refresh window, combining real-time driving state and road condition information, and adopting a multi-stage power scheduling and optimization method, it realizes the coordinated driving of multiple motors of an electric tricycle. After the driving state is recognized, power distribution analysis and optimization are carried out, and the distributed driving power is replaced by smooth transition calculation to ensure stable driving performance under different road conditions, achieving the technical effects of improving the accuracy of motor power distribution and enhancing driving stability.

[0012] Next, the technical solutions in the embodiments of this application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all the embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of this application.

[0013] It should be noted that any variations of the terms "including" and "having" are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or server that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or are inherent to these processes, methods, products, or devices.

[0014] Embodiment, as Figure 1 shown, the embodiment of this application provides a motor power scheduling system combined with dynamic feedback, and this system includes: A predefined module 11, and the predefined module 11 is used to predefine a driving state refresh window and a driving road condition refresh window.

[0015] In the embodiment of this application, first, a driving state refresh window and a driving road condition refresh window are predefined through the predefined module. Among them, the driving state refresh window is set to 1 to 2 seconds. This time interval means that the driving state of the electric tricycle, such as acceleration, braking, turning, etc., is recognized and updated every 1 to 2 seconds. The driving road condition refresh window is set to 3 to 5 seconds. This time interval is used to regularly update the road condition information of the electric tricycle, such as whether the road surface is flat, slope change, whether there are obstacles, etc.

[0016] A driving state recognition module 12, and the driving state recognition module 12 is used to perform the driving state recognition of the electric tricycle with the driving state refresh window as a constraint to obtain real-time driving state information.

[0017] In the embodiment of the present application, the driving state recognition module recognizes the driving state of the electric tricycle in real time according to a predefined driving state refresh window, and obtains corresponding real-time driving state information, including real-time driving acceleration, real-time driving steering angle, real-time driving slope, real-time driving wheel speed difference, and real-time driving speed, etc.

[0018] Specifically, the real-time driving acceleration is collected by an acceleration sensor on the electric tricycle. The acceleration sensor is a three-axis accelerometer based on MEMS (Micro-Electro-Mechanical System) technology, which simultaneously measures the acceleration changes of the vehicle along the X-axis, Y-axis, and Z-axis. By measuring the acceleration changes of the vehicle, it is judged whether the vehicle is accelerating, decelerating, or maintaining a constant speed. The real-time driving steering angle is obtained by a gyroscope and a steering wheel angle sensor. The gyroscope is mainly used to measure the rotation rate of the vehicle, while the steering wheel angle sensor directly measures the rotation angle of the steering wheel. Through these two sensors, the steering angle of the vehicle and whether the vehicle is turning or changing lanes are judged. The real-time driving slope information is obtained by an inclination sensor or a GPS sensor. During the driving of the vehicle, the inclination sensor measures the inclination angle of the vehicle relative to the horizontal plane, thereby judging the slope state of the vehicle. The real-time driving wheel speed difference is collected by wheel speed sensors on the wheels. Each wheel motor is equipped with a wheel speed sensor, and these sensors monitor the rotational speed and wheel speed of each motor. The wheel speed difference is the wheel speed difference between the two rear wheels of the tricycle. When the rotational speed of one wheel is different from that of the other rear wheel, a wheel speed difference will occur, which usually reflects whether the vehicle has skidding or uneven power. The real-time driving speed is obtained by a vehicle speed sensor, such as GPS.

[0019] The power distribution analysis module 13, and the power distribution analysis module 13 is used to perform power distribution analysis according to the real-time driving state information and output distributed drive power.

[0020] Further, as Figure 2 shown, in the system provided by the embodiment of the application, the power distribution analysis module 13 is further used for: Defining a previous data collection starting point according to the output node of the real-time driving state information; extracting previous driving state information based on the driving state refresh window and the previous data collection starting point, where the previous driving state information includes previous driving acceleration, previous driving steering angle, previous driving slope, previous driving wheel speed difference, and previous driving speed; performing state transition judgment based on the real-time driving state information and the previous driving state information, and generating the distributed drive power according to the judgment result.

[0021] In the embodiment of the present application, the power distribution analysis module first defines a previous data collection starting point according to the output node of the real-time driving state information. Specifically, the time node when the real-time driving state information is obtained is used as the previous data collection starting point.

[0022] Next, based on the driving state refresh window and the previous data collection starting point, the previous driving state information is extracted. Specifically, based on the driving state refresh window, trace back to the time node of the previous driving state refresh window before the previous data collection starting point. Here, the driving state refresh window represents a predetermined time length. Within this time window, trace back to a certain moment before the output node to extract the corresponding previous driving state information. The extracted previous driving state information includes the previous driving acceleration, the previous driving steering angle, the previous driving slope, the previous driving wheel speed difference, and the previous driving speed, which are obtained by real-time monitoring in historical time.

[0023] Finally, based on the real-time driving state information and the previous driving state information, a state transition judgment is made. The similarity between the real-time driving state information and the previous driving state information is judged. If these two pieces of information are similar, the previous driving state information in the previous driving state information is extracted, and the previous driving state information is used as the distributed drive power output. If the similarity between the real-time driving state information and the previous driving state information is less than a preset value, the real-time driving state information is loaded into a pre-trained power scheduling analysis model to output an updated drive power. Then, a smooth transition calculation is performed on the updated drive power and the previous drive power to generate the distributed drive power.

[0024] Furthermore, in the system provided by the application embodiment, the power distribution analysis module 13 is further configured to: Calculate the Pearson correlation coefficient according to the index composition of the real-time driving state information and the previous driving state information to obtain a state consistency coefficient; if the state consistency coefficient is less than the preset value Q, load the real-time driving state information into the power scheduling analysis model to output an updated drive power; load the previous drive power according to the previous driving state information, and perform a smooth transition calculation according to the updated drive power and the previous drive power to generate the distributed drive power; if the state consistency coefficient is greater than the preset value Q, load the previous drive power according to the previous driving state information, and output the previous drive power as the distributed drive power.

[0025] In the embodiment of the present application, first, the Pearson correlation coefficient is calculated according to the index composition of the real-time driving state information and the previous driving state information to obtain a state consistency coefficient.

[0026] Next, the calculated state consistency coefficient is compared with the preset value Q. If the state consistency coefficient is less than the preset value Q, it is considered that the similarity between the real-time driving state information and the previous driving state information is small, and the real-time driving state information is loaded into the power scheduling analysis model to output an updated drive power. Among them, the preset value Q is preset and is a value greater than 0 and less than 1, which is set by technical experts according to actual needs.

[0027] Next, the previous driving power is loaded according to the previous driving state information, and a smooth transition calculation is performed based on the updated driving power and the previous driving power to generate a distributed driving power. Specifically, first, a preset smoothing coefficient is obtained. By calculating the difference between 1 and the smoothing coefficient, multiplying it by the previous driving power, and then adding the smoothing coefficient multiplied by the updated driving power, the distributed driving power is obtained.

[0028] If the calculated state consistency coefficient is greater than the preset value Q, it indicates that the difference between the current driving state and the previous driving state is small. At this time, instead of updating the driving power, the previous driving power is directly used as the output of the current distributed driving power.

[0029] Furthermore, in the system provided by the application embodiment, the power distribution analysis module 13 is further configured to: Interactively obtain M sample state feature information, M sample driving speed sets, and M sample power distribution data pairs of M sample driving states; construct a driving state recognition layer based on the M sample driving states and M sample state feature information; perform multivariate function fitting on the M sample driving speed sets and M sample power distribution data pairs, and output M power distribution calculation functions; construct M power distribution calculation branches based on the M power distribution calculation functions, and complete the construction of the driving speed calculation layer by connecting the power distribution calculation branches in parallel; connect the output end of the driving state recognition layer to the input end of the driving speed calculation layer to complete the construction of the power scheduling analysis model.

[0030] In the embodiment of the present application, M sample state feature information, M sample driving speed sets, and M sample power distribution data pairs of M sample driving states are obtained through an interactive historical database. These samples correspond one by one, and one sample state feature information corresponds to one sample driving speed set and one sample power distribution data pair. Among them, the sample driving state refers to the state of the vehicle in different driving scenarios, such as straight driving, acceleration, deceleration, climbing, etc.; the sample state feature information includes features such as the vehicle's acceleration, steering angle, slope, wheel speed difference, and speed; the sample driving speed set reflects the actual driving speed under different samples, and the sample power distribution data pair refers to the motor power distribution data related to these driving states.

[0031] Next, based on the obtained M sample driving states and M sample state feature information, a driving state recognition layer is constructed. Specifically, using the method of supervised learning, a suitable classification algorithm, such as a support vector machine, is selected to classify the driving states. The M sample driving states are used as labels, and the M sample state feature information is used as input features for training to obtain a driving state recognition layer. The driving state recognition layer can judge the current driving state of the vehicle according to the real-time collected vehicle data (such as acceleration, steering angle, etc.), for example, judge whether the vehicle is in an accelerating, turning or decelerating state, etc.

[0032] After the driving state recognition layer is completed, based on the M sample driving speed sets and M sample power distribution data pairs, a driving speed calculation layer is constructed. This layer is mainly realized by fitting a multivariable function of the sample driving speed and power distribution data. By using multivariable regression analysis (such as polynomial regression or linear regression), the relationship between the driving speed and power distribution is modeled. The goal of this process is to deduce the corresponding driving speed from various driving state features (such as vehicle acceleration, steering angle, etc.). The fitted function can calculate the driving speed according to the real-time parameters of the vehicle (such as acceleration, steering angle, etc.) in different states, so as to provide data support for power scheduling. Through multivariable regression analysis, M power distribution calculation functions are obtained. The M power distribution calculation functions respectively correspond to M different driving states, and each function is specifically used to calculate the power output matching that state.

[0033] Next, based on the obtained M power distribution calculation functions, M power distribution calculation branches are constructed, and each branch corresponds to the power distribution rule in one driving state.

[0034] Then, by connecting the M power distribution calculation branches in parallel, the construction of the driving speed calculation layer is completed. This calculation layer can calculate the driving power in real time according to the state feature information of the vehicle in different driving states.

[0035] Finally, by connecting the output end of the driving state recognition layer to the input end of the driving speed calculation layer, the construction of the power scheduling analysis model is completed.

[0036] Furthermore, in the system provided by the application embodiment, the real-time driving state information includes real-time driving acceleration, real-time driving steering angle, real-time driving slope, real-time driving wheel speed difference and real-time driving speed.

[0037] In the embodiment of the present application, the real-time driving state information includes real-time driving acceleration, real-time driving steering angle, real-time driving slope, real-time driving wheel speed difference and real-time driving speed. At the same time, the real-time driving state information also includes driving power, and the driving power refers to the power output by each motor at a certain moment. The power output of the motor is monitored in real time through a power sensor.

[0038] Further, in the system provided by the application embodiment, the power distribution analysis module 13 is further configured to: Load the real-time driving acceleration, real-time driving steering angle, real-time driving slope, and real-time driving wheel speed difference into the driving state recognition layer of the power scheduling analysis model for driving state recognition to obtain the real-time driving state type; traverse the M power distribution calculation branches in the driving speed calculation layer by using the real-time driving state type to activate the real-time distribution calculation branch; load the real-time driving speed into the distribution calculation branch to calculate and obtain the updated driving power.

[0039] In the embodiment of the present application, first, the foregoing obtained real-time driving acceleration, real-time driving steering angle, real-time driving slope, and real-time driving wheel speed difference are loaded into the driving state recognition layer of the constructed power scheduling analysis model for driving state recognition to obtain the real-time driving state type.

[0040] Once the real-time driving state type is recognized, use this type to traverse the M power distribution calculation branches in the driving speed calculation layer. Each calculation branch corresponds to a specific driving state (such as acceleration, climbing, etc.) and includes a power distribution calculation function related thereto. Activate the calculation branch matching the current driving state type, and on this basis, load the real-time driving speed data into the activated power distribution calculation branch for power calculation. These calculations will consider the current speed of the vehicle and the recognized driving state, and calculate the updated driving power through the corresponding power distribution calculation function. This driving power represents the power value that each motor of the electric tricycle should output under the current driving state.

[0041] A driving road condition refresh window activation module 14, where the driving road condition refresh window activation module 14 is configured to activate the driving road condition refresh window according to the generation node of the distributed driving power when performing multi-motor coordinated driving of the electric tricycle by using the distributed driving power.

[0042] In the embodiment of the present application, the driving road condition refresh window activation module activates the corresponding driving road condition refresh window according to the generation node of the distributed driving power of the electric tricycle, so as to obtain the latest information related to the current road condition, thereby optimizing power scheduling and motor coordination.

[0043] Specifically, when performing multi-motor coordinated driving of the electric tricycle by using the distributed driving power, the generation node of the distributed driving power is used as a trigger signal to activate the corresponding driving road condition refresh window. At this time, the driving road condition refresh window performs data collection at a predefined time interval based on the generation node of the distributed driving power.

[0044] The driving road condition acquisition module 15 is configured to acquire the driving road conditions of the electric tricycle with the driving road condition refresh window as a constraint, so as to obtain real-time driving road condition information.

[0045] In the embodiment of the present application, the driving road condition acquisition module acquires the driving road conditions of the electric tricycle with the driving road condition refresh window as a constraint. The driving road condition refresh window is a time-driven window that specifies the timing of acquiring road condition data. Whenever the distributed driving power of the electric tricycle is updated, the driving road condition refresh window is activated, and after a preset time length of the driving road condition refresh window, the driving road conditions of the electric tricycle are acquired.

[0046] Specifically, the driving road condition acquisition module obtains the real-time friction coefficient through in-vehicle sensors and external devices, which can be deduced through ground sensors or slip ratio detectors; the real-time road surface unevenness is analyzed through acceleration sensors and wheel vibration data; the road surface type (such as asphalt, dirt, etc.) is obtained through the in-vehicle imaging system or ground radar detection technology; the real-time temperature and weather information are obtained through in-vehicle meteorological monitoring instruments and by connecting to external weather data sources (such as meteorological APIs). Through this process, real-time driving road condition information is obtained.

[0047] The power distribution optimization module 16 is configured to optimize the power distribution of the distributed driving power according to the real-time driving road condition information, so as to obtain an optimized power distribution.

[0048] Furthermore, in the system provided by the embodiment of the application, the power distribution optimization module 16 is further configured to: By analyzing the real-time driving road condition information, obtain the real-time friction coefficient, real-time road surface unevenness, real-time road surface type, real-time temperature and real-time rainfall; generate a dynamic road model based on the real-time driving slope, real-time friction coefficient, real-time road surface unevenness and real-time road surface type; perform power summation based on the distributed driving power to calculate the total driving power; starting from the distributed driving power and with the total driving power as the power conservation constraint, randomly generate N power distribution schemes; in the dynamic road model, use the real-time driving acceleration, real-time driving steering angle, real-time driving wheel speed difference and real-time driving speed to simulate the driving control process of the electric tricycle, and use the N power distribution schemes to simulate the motor power scheduling process of the electric tricycle, and output N groups of simulated driving performance information, where each group of simulated driving performance information includes driving energy consumption, driving force distribution efficiency, trajectory deviation degree and slip ratio; according to the N groups of simulated driving performance information, perform iterative simulation optimization of the N power distribution schemes in the dynamic road model, and output the optimized power distribution.

[0049] In the embodiments of the present application, the power distribution optimization module obtains key information including real-time friction coefficient, real-time road surface unevenness, real-time road surface type, real-time temperature, and real-time rainfall by analyzing real-time driving road conditions information. These data are collected through in-vehicle sensors and external data sources. Specifically, the friction coefficient is usually obtained through slip sensors in contact with the road surface by the wheels, the road surface unevenness and type are detected by lidar or computer vision sensors (such as stereo vision cameras), and the temperature and rainfall are obtained by in-vehicle meteorological sensors or networked meteorological services.

[0050] Then, based on the above real-time road conditions information, the power distribution optimization module uses regression analysis methods to fit a dynamic road model. Specifically, by using algorithms such as least squares regression or other fitting algorithms, factors such as road surface slope, friction coefficient, road surface unevenness, and road surface type are transformed into a mathematical model. This process generates a dynamic road model by fitting real-time road conditions data, which is used to reflect the driving performance of the electric tricycle under different road conditions. For example, the greater the road surface unevenness and the lower the friction coefficient, the greater the driving force and energy consumption required by the vehicle.

[0051] Next, the power distribution optimization module sums up the current distributed drive powers to calculate the total drive power. This total power is the overall power required by the electric tricycle, and the sum of the powers of all motors is the total drive power. At this time, the obtained total drive power serves as a constraint condition for power conservation, providing a basis for the optimization calculation.

[0052] After determining the total drive power, N power distribution schemes are randomly generated as the power distribution methods for different motors. To generate these power distribution schemes, methods based on genetic algorithms or particle swarm optimization are used to simulate multiple possible power distribution strategies to find the optimal scheme. These randomly generated power distribution schemes will be used in the following simulations to evaluate their performance in actual driving.

[0053] Then, the power distribution optimization module uses the dynamic road model and real-time driving data (such as acceleration, steering angle, wheel speed difference, speed, etc.) to simulate the above N power distribution schemes, simulating the actual driving process of the electric tricycle. These simulation processes are mainly completed through numerical calculations and simulation models. For example, the finite difference method or a simulation tool based on vehicle dynamics is used to predict the driving of the electric tricycle. The simulation results generate N sets of driving performance data, and each set of data includes key performance indicators such as driving energy consumption, driving force distribution efficiency, trajectory deviation degree, and slip rate.

[0054] The final power distribution optimization module performs iterative optimization based on N groups of simulated driving performance information to screen out the optimal power distribution scheme. The iterative process is based on the optimization objectives of the simulation results, such as the best output performance stability, and uses genetic algorithms or other heuristic optimization methods to optimize the power distribution. After several iterations, the final optimized power distribution scheme is output to obtain the optimal distributed drive power.

[0055] Further, in the system provided by the application embodiment, the power distribution optimization module 16 is further configured to: Evaluate the N groups of simulated driving performance information based on the driving performance weight configuration, and output N driving performance stability degrees; extract the first screening distribution scheme and the second screening distribution scheme from the N power distribution schemes by serializing the N driving performance stability degrees; use the total drive power as the power conservation constraint, perform crossover mutation on the first screening distribution scheme and the second screening distribution scheme to obtain the first offspring distribution scheme and the second offspring distribution scheme; perform random perturbation of the power distribution on the first offspring distribution scheme and the second offspring distribution scheme to generate the first group of updated distribution schemes and the second group of updated distribution schemes; simulate the motor power scheduling process of the electric tricycle using the first offspring distribution scheme, the second offspring distribution scheme, the first group of updated distribution schemes and the second group of updated distribution schemes, and perform driving performance evaluation using the driving performance weight configuration to output the first offspring performance stability degree, the second offspring performance stability degree, the first group of updated performance stability degrees and the second group of updated performance stability degrees; locate the third offspring distribution scheme and the fourth offspring distribution scheme from the first offspring distribution scheme, the second offspring distribution scheme, the first group of updated distribution schemes and the second group of updated distribution schemes according to the first offspring performance stability degree, the second offspring performance stability degree, the first group of updated performance stability degrees and the second group of updated performance stability degrees; and so on, perform crossover mutation and random perturbation of the power distribution of the offspring distribution scheme based on the performance stability degree until the preset number of iterations is reached, and output the optimized power distribution with the minimum performance stability degree.

[0056] In the embodiments of the present application, the driving performance weight configuration, as a preset parameter set, is used to assign weights to different driving performance indicators (such as driving energy consumption, driving force distribution efficiency, trajectory deviation degree, and slip rate). Each indicator is assigned a weight value according to its importance in vehicle control. These weight configurations are preset by technical experts. Next, N groups of simulated driving performance information (including driving energy consumption, driving force distribution efficiency, trajectory deviation degree, and slip rate) are normalized. The purpose of the normalization operation is to eliminate the dimensional differences between different performance indicators, so that they are within the same scale range, facilitating comprehensive evaluation and comparison. Common normalization methods include linear normalization (for example, mapping each indicator value to between 0 and 1) or Z-score standardization (transforming according to the mean and standard deviation of each indicator). Through normalization, all driving performance data is converted into standardized numerical values, eliminating the problem of different dimensions, so that each indicator can be compared under the same conditions. After completing the normalization process, the driving performance weight configuration is used to perform weighted summation on these N groups of normalized simulated driving performance information to obtain N driving performance stability degrees.

[0057] Then, through a sorting algorithm (such as quicksort or mergesort), the N driving performance stability degrees are sorted, and the top two best-performing schemes are selected, called the first screening and allocation scheme and the second screening and allocation scheme.

[0058] After screening out the first screening and allocation scheme and the second screening and allocation scheme, based on the power conservation constraint (that is, the total driving power remains unchanged), a crossover and mutation operation is performed on them. The crossover and mutation simulates the crossover operation in the genetic algorithm. By exchanging some power distribution parameters in these two groups of schemes, new power distribution schemes are generated, namely the first offspring allocation scheme and the second offspring allocation scheme. These newly generated schemes provide diverse choices for further optimization.

[0059] Then, a random perturbation of the power distribution is performed on the offspring schemes after crossover and mutation, that is, a small adjustment within a preset range is made to the power distribution parameters of each scheme to obtain the first group of updated allocation schemes and the second group of updated allocation schemes.

[0060] On this basis, the first offspring allocation scheme, the second offspring allocation scheme, the first group of updated allocation schemes, and the second group of updated allocation schemes are used to simulate the motor power scheduling process of the electric tricycle, and the driving performance weight configuration is used for driving performance evaluation, and the first offspring performance stability degree, the second offspring performance stability degree, the first group of updated performance stability degree, and the second group of updated performance stability degree are output.

[0061] Subsequently, according to the calculated first-generation performance stability, second-generation performance stability, first-group update performance stability, and second-group update performance stability, select the two allocation schemes with the top two stabilities from the first-generation allocation scheme, second-generation allocation scheme, first-group update allocation scheme, and second-group update allocation scheme as the third-generation allocation scheme and the fourth-generation allocation scheme.

[0062] After that, continue with the crossover mutation and random perturbation of the power allocation in the offspring allocation scheme based on the performance stability, and iterate and optimize until the preset number of iterations is reached. During each iteration process, the power allocation scheme is continuously optimized, and the finally output optimized power allocation is the scheme with the minimum driving performance stability, so as to ensure the optimal power scheduling of the electric tricycle under the current road conditions.

[0063] Through the above steps, the power distribution optimization module finally obtains the optimized power allocation according to the real-time road conditions and driving requirements, using a variety of optimization algorithms (including normalization, regression analysis, crossover mutation, and random perturbation, etc.), to ensure that the electric tricycle can provide the best power output, optimal energy efficiency, and best driving experience under different driving conditions.

[0064] The multi-motor coordinated drive module 17, where the multi-motor coordinated drive module 17 is used to smoothly replace the distributed drive power with the optimized power allocation for the multi-motor coordinated drive of the electric tricycle.

[0065] In the embodiment of the present application, the multi-motor coordinated drive module coordinates and controls the three motors of the tricycle by adopting the optimized power allocation scheme. In the multi-motor system of the electric tricycle, each motor is responsible for different driving tasks, such as driving the front wheel, rear wheel, or wheels on both sides. Through coordinated control, each motor can adjust its working state according to the optimized power allocation scheme to achieve smooth acceleration, deceleration, or steering of the vehicle.

[0066] In the embodiment of the present application, in summary, the embodiment of the present application has at least the following technical effects: This application predefines a driving state refresh window and a driving road condition refresh window; constrained by the driving state refresh window, the driving state of the electric tricycle is recognized to obtain real-time driving state information; power distribution analysis is performed based on the real-time driving state information to output distributed drive power; when the distributed drive power is used for the multi-motor coordinated drive of the electric tricycle, the driving road condition refresh window is activated according to the generation node of the distributed drive power; constrained by the driving road condition refresh window, the driving road conditions of the electric tricycle are collected to obtain real-time driving road condition information; the distributed drive power is optimized for power distribution according to the real-time driving road condition information to obtain optimized power distribution; the optimized power distribution is used to smoothly replace the distributed drive power for the multi-motor coordinated drive of the electric tricycle. The present invention solves the technical problem in the prior art that the inaccurate motor power distribution leads to poor driving stability. By predefining the driving state refresh window and the driving road condition refresh window, combining real-time driving state and road condition information, and adopting a multi-stage power scheduling and optimization method, the multi-motor coordinated drive of the electric tricycle is realized. After the driving state is recognized, power distribution analysis and optimization are carried out, and the distributed drive power is replaced by smooth transition calculation to ensure smooth driving performance under different road conditions, achieving the technical effects of improving the accuracy of motor power distribution and enhancing driving stability.

[0067] It should be noted that the above sequence of embodiments of the present application is only for description and does not represent the superiority or inferiority of the embodiments. And the above specific embodiments of this specification have been described. The processes depicted in the drawings do not necessarily require the specific order and continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0068] The above are only the preferred embodiments of the present application and are not intended to limit the present application. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application shall be included within the protection scope of the present application.

[0069] This specification and the drawings are only exemplary descriptions of the present application and are considered to have covered any and all modifications, variations, combinations, or equivalents within the scope of the present application. Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalent technologies, the present application is intended to include these changes and modifications.

Claims

1. A motor power dispatching system combined with dynamic feedback, characterized in that: The system comprises: A predefined module, the predefined module is used to predefine a driving status refresh window and a driving road condition refresh window; A driving state recognition module, the driving state recognition module is used to recognize the driving state of the electric tricycle with the driving state refresh window as a constraint to obtain real-time driving state information; A power distribution analysis module, the power distribution analysis module is used to perform power distribution analysis according to the real-time driving state information and output distributed driving power; A driving condition refresh window activation module, wherein the driving condition refresh window activation module is used to activate the driving condition refresh window according to a generation node of the distributed driving power when the distributed driving power is used to coordinately drive the multi-motor of the electric tricycle; A driving road condition collection module, the driving road condition collection module is used to collect the driving road condition of the electric tricycle with the driving road condition refresh window as a constraint to obtain real-time driving road condition information; A power distribution optimization module, the power distribution optimization module is used to optimize the power distribution of the distributed driving power according to the real-time driving road condition information to obtain an optimized power distribution; A multi-motor coordinated driving module is used to use the optimized power distribution to smoothly replace the distributed driving power to perform multi-motor coordinated driving of the electric tricycle.

2. The motor power dispatching system combined with dynamic feedback according to claim 1, characterized in that: The power distribution analysis module is also used for: Defining a starting point for collecting preceding data according to an output node of the real-time driving status information; Extracting the preceding driving state information according to the driving state refresh window and the preceding data collection starting point, wherein the preceding driving state information includes the preceding driving acceleration, the preceding driving steering angle, the preceding driving slope, the preceding driving wheel speed difference and the preceding driving speed; A state transition judgment is performed based on the real-time driving state information and the previous driving state information, and the distributed driving power is generated according to the judgment result.

3. The motor power dispatching system combined with dynamic feedback as claimed in claim 2, characterized in that: The power allocation analysis module is further used for: Calculate the Pearson correlation coefficient based on the index composition of the real-time driving state information and the previous driving state information to obtain a state consistency coefficient; If the state consistency coefficient is less than a preset value Q, the real-time driving state information is loaded into a power scheduling analysis model, and the updated driving power is output; Loading the previous driving power according to the previous driving state information, and performing smooth transition calculation according to the updated driving power and the previous driving power to generate the distributed driving power; If the state consistency coefficient is greater than a preset value Q, the preceding driving power is loaded according to the preceding driving state information, and the preceding driving power is output as the distributed driving power.

4. The motor power dispatching system combined with dynamic feedback as claimed in claim 3, characterized in that: The power distribution analysis module is also used for: Interactively obtain M sample state feature information of M sample driving states, M sample driving speed sets and M sample power allocation data pairs; Constructing a driving state recognition layer based on the M sample driving states and the M sample state feature information; Performing multivariable function fitting on the M sample driving speed sets and the M sample power allocation data pairs, and outputting M power allocation calculation functions; Constructing M power allocation calculation branches based on the M power allocation calculation functions, and completing the construction of the driving speed calculation layer by connecting the power allocation calculation branches in parallel; The output end of the driving state recognition layer is connected to the input end of the driving speed calculation layer to complete the construction of the power scheduling analysis model.

5. The motor power dispatching system combined with dynamic feedback as claimed in claim 4, characterized in that: The real-time driving state information includes real-time driving acceleration, real-time driving steering angle, real-time driving slope, real-time driving wheel speed difference and real-time driving speed.

6. The motor power dispatching system combined with dynamic feedback as claimed in claim 5, characterized in that: The power distribution analysis module is also used for: Loading the real-time driving acceleration, real-time driving steering angle, real-time driving slope and real-time driving wheel speed difference into the driving state recognition layer of the power scheduling analysis model to perform driving state recognition and obtain the real-time driving state type; Using the real-time driving state type to traverse the M power allocation calculation branches in the driving speed calculation layer, activating the real-time allocation calculation branch; The real-time driving speed is loaded into the distribution calculation branch to calculate and obtain the updated driving power.

7. The motor power dispatching system combined with dynamic feedback as claimed in claim 5, characterized in that: The power distribution optimization module is used to: By analyzing the real-time driving road condition information, a real-time friction coefficient, a real-time road surface roughness, a real-time road surface type, a real-time temperature and a real-time rainfall amount are obtained; Generate a dynamic road model based on the real-time driving slope, real-time friction coefficient, real-time road surface roughness and real-time road surface type fitting; Perform power addition based on the distributed drive power to calculate the total drive power; Taking the distributed drive power as a starting point and the total drive power as a power conservation constraint, N power allocation schemes are randomly generated; In the dynamic road model, the real-time driving acceleration, real-time driving steering angle, real-time driving wheel speed difference and real-time driving speed are used to simulate the driving control process of the electric tricycle, and the N power allocation schemes are used to simulate the motor power scheduling process of the electric tricycle, and N groups of simulated driving performance information are output, wherein each group of simulated driving performance information includes driving energy consumption, driving force allocation efficiency, trajectory deviation and slip rate; According to the N groups of simulated driving performance information, iterative simulation optimization of the N power distribution schemes is performed in the dynamic road model, and the optimized power distribution is output.

8. The motor power dispatching system combined with dynamic feedback as claimed in claim 7, characterized in that: The power distribution optimization module is used to: Evaluating the N sets of simulated driving performance information based on the driving performance weight configuration, and outputting N driving performance stability levels; By serializing the N driving performance stabilities, a first screening allocation scheme and a second screening allocation scheme are extracted from the N power allocation schemes; Taking the total driving power as a power conservation constraint, cross-mutating the first screening allocation scheme and the second screening allocation scheme to obtain a first offspring allocation scheme and a second offspring allocation scheme; Performing random perturbations of power allocation on the first child allocation scheme and the second child allocation scheme to generate a first group of updated allocation schemes and a second group of updated allocation schemes; The first child generation allocation scheme, the second child generation allocation scheme, the first group update allocation scheme and the second group update allocation scheme are used to simulate the motor power scheduling process of the electric tricycle, and the driving performance weight configuration is used to evaluate the driving performance, and the first child generation performance stability, the second child generation performance stability, the first group update performance stability and the second group update performance stability are output; Locating a third child allocation scheme and a fourth child allocation scheme from the first child allocation scheme, the second child allocation scheme, the first group of updated allocation schemes, and the second group of updated allocation schemes according to the first child performance stability, the second child performance stability, the first group of updated performance stability, and the second group of updated performance stability; By analogy, cross-mutation of the offspring allocation scheme and random perturbation of the power allocation are performed based on the performance stability until a preset number of iterations is reached, and the optimized power allocation with the minimum performance stability is output.

Images