Vehicle transmission brake control method and device and vehicle

Abstract

The application provides a vehicle gearbox brake control method and device and a vehicle. The method comprises the following steps: acquiring current state information in real time, wherein the current state information comprises an input shaft rotating speed, an output shaft rotating speed, a gearbox body temperature, a current gear of the gearbox, a target gear of the gearbox and an intermediate shaft brake slope; inputting the current state information into an optimized network model to obtain control time information, wherein the control time information comprises a brake start time and a brake total time; the optimized network model is obtained by training and optimizing an original network model, wherein the original network model is trained by taking historical state information as input and taking historical control time information as target output; and generating a control instruction according to the brake start time and the brake total time and sending the control instruction to the intermediate shaft brake. The method is beneficial to improving the brake accuracy of the intermediate shaft brake.

Description

Technical Field

[0001] This application relates to braking control technology, and more particularly to a vehicle transmission braking control method, device, and vehicle. Background Technology

[0002] The current control strategy of the intermediate shaft brake assist device in automatic transmissions is based on the speed difference between the engagement sleeve and the gear to be engaged during the shifting process, controlling the action of the intermediate shaft brake. Specifically, when a shift is detected during driving, the transmission controller sends a control command to open the solenoid valve of the intermediate shaft brake's air inlet. Compressed gas enters the working cylinder inside the brake, and the resulting pressure overcomes the resistance of the return spring, rapidly squeezing the friction pads. The friction pads interact with the intermediate shaft, causing the intermediate shaft speed to decrease. When the detected speed difference reaches a certain value, the transmission controller sends a control signal to close the air inlet solenoid valve and open the exhaust solenoid valve. The friction pads, along with the return spring, expel the high-pressure gas in the cylinder to the atmosphere. The intermediate shaft brake and the intermediate shaft gradually lose their force and no longer generate friction, thus completing the intermediate shaft braking.

[0003] Currently, most intermediate shaft brakes use fixed parameters to control their operation, resulting in poor braking accuracy and affecting shifting quality. Therefore, there is an urgent need to design a new control method for intermediate shaft brakes to improve the accuracy of their operating time. Summary of the Invention

[0004] This application provides a vehicle transmission braking control method, device, and vehicle, which provides a highly accurate transmission braking control method.

[0005] On one hand, this application provides a vehicle transmission braking control method, executed by the transmission brake, the method comprising:

[0006] Real-time acquisition of current status information, including input shaft speed, output shaft speed, gearbox housing temperature, current gear of the gearbox, target gear of the gearbox, and slope of the intermediate shaft brake;

[0007] The current state information is input into the optimized network model to obtain control time information, which includes braking start time and total braking time. The optimized network model is obtained by training and optimizing the original network model. During training, the original network model takes historical state information as input and historical control time information as the target output.

[0008] Based on the braking start time and the total braking time, a control command is generated and sent to the intermediate shaft brake. The control command instructs the intermediate shaft brake to start braking at the braking start time and to stop braking when the total braking time is reached.

[0009] In another possible implementation, before inputting the current state information into the optimized network model, the method further includes:

[0010] Acquire training data, which includes training samples, and the training samples include historical state information and historical control time information corresponding to the historical state information;

[0011] Take integer values ​​for Y within the range of 1 to a preset number of iterations, and perform the following steps until the error of the Yth iteration satisfies the termination training condition, and / or Y is equal to the preset number of iterations, to obtain the trained network model; wherein, the termination training condition is: the error of the Yth iteration is within the preset range;

[0012] The historical state information is input into the original network model to obtain the Y-th prediction control time information;

[0013] The Y-th error is calculated based on the Y-th predicted control time information and the historical control time information.

[0014] Based on the Y-th error, determine the updated model weights and the updated model threshold, and update the original network model based on the updated model weights and the updated model threshold.

[0015] In another possible implementation, before inputting the historical control time information into the original network model, the method further includes:

[0016] Based on the training data, the topology of the original network model is obtained; the topology includes an input layer with 6 neurons, a hidden layer with m neurons, and an output layer with 2 neurons.

[0017] Based on the topology, the number of model weights and model thresholds is determined, and the model weights and model thresholds are initialized to obtain initial weights and initial thresholds, thereby obtaining the original network model.

[0018] In another possible implementation, before inputting the current state information into the optimized network model, the method further includes:

[0019] Based on the trained network model, a preset number of chromosomes are randomly generated to form a population; the chromosomes serve as the updated model weights and the updated model thresholds of the trained network model.

[0020] Within the range of 1 to the preset number of iterations, take integer values ​​for X in sequence and perform the following steps until X equals the preset number of iterations. Then, determine the optimal individual based on the current population and update the update model weights and the update model threshold based on the optimal individual to obtain the optimized network model.

[0021] The historical state information is input into the trained network model to obtain the Xth prediction control time information;

[0022] The fitness of the Xth individual is calculated based on the predicted control time information of the Xth time and the historical control time information corresponding to the historical state information.

[0023] The current population is determined based on the fitness of the Xth individual.

[0024] In another possible implementation, determining the current population based on the fitness of the Xth individual includes:

[0025] A new population is formed by selecting chromosomes with high fitness from X individuals according to a preset algorithm; the preset algorithm is any one of roulette wheel selection, tournament selection, and competitive selection.

[0026] The chromosomes in the new population are subjected to crossover and mutation operations to obtain the current population.

[0027] In another possible implementation, the training data further includes test samples, and after obtaining the optimized network model, the method further includes:

[0028] The optimized network model is tested based on the test samples to obtain test results;

[0029] If the test results indicate that the optimized network model does not meet the preset accuracy requirements, then the steps for obtaining the trained network model are repeated.

[0030] In another possible implementation, acquiring the training data includes:

[0031] Acquire historical state information and historical control time information of the transmission during upshifting under different operating conditions within a preset time period, and use the historical state information and the historical control time information as sample data;

[0032] The sample data is preprocessed to obtain the training data. The preprocessing includes filtering ideal sample data from the sample data and normalizing the ideal sample data. The historical control time information of the ideal sample data is within a preset reasonable range.

[0033] The training data is divided into training samples and test samples according to a preset ratio. The training samples are used to train the original network model, and the test samples are used to test the optimized network model.

[0034] Secondly, this application provides a vehicle transmission braking control device, comprising:

[0035] The acquisition module is used to acquire current status information in real time, including input shaft speed, output shaft speed, gearbox housing temperature, current gearbox gear, target gearbox gear, and intermediate shaft brake slope.

[0036] The prediction module is used to input the current state information into the optimized network model to obtain control time information, which includes braking start time and total braking time. The optimized network model is obtained by training and optimizing the original network model. During training, the original network model takes historical state information as input and historical control time information as the target output.

[0037] The control module is used to generate a control command based on the braking start time and the total braking time, and send the control command to the intermediate shaft brake. The control command is used to instruct the intermediate shaft brake to start braking at the braking start time and to stop braking when the total braking time is reached.

[0038] In another possible implementation, the prediction module is also used for:

[0039] Acquire training data, which includes training samples, and the training samples include historical state information and historical control time information corresponding to the historical state information;

[0040] Take integer values ​​for Y within the range of 1 to a preset number of iterations, and perform the following steps until the error of the Yth iteration satisfies the termination training condition, and / or Y is equal to the preset number of iterations, to obtain the trained network model; wherein, the termination training condition is: the error of the Yth iteration is within the preset range;

[0041] The historical state information is input into the original network model to obtain the Y-th prediction control time information;

[0042] The Y-th error is calculated based on the Y-th predicted control time information and the historical control time information.

[0043] Based on the Y-th error, determine the updated model weights and the updated model threshold, and update the original network model based on the updated model weights and the updated model threshold.

[0044] In another possible implementation, the prediction module is also used for:

[0045] Based on the training data, the topology of the original network model is obtained; the topology includes an input layer with 6 neurons, a hidden layer with m neurons, and an output layer with 2 neurons.

[0046] Based on the topology, the number of model weights and model thresholds is determined, and the model weights and model thresholds are initialized to obtain initial weights and initial thresholds, thereby obtaining the original network model.

[0047] In another possible implementation, the prediction module is also used for:

[0048] Based on the trained network model, a preset number of chromosomes are randomly generated to form a population; the chromosomes serve as the updated model weights and the updated model thresholds of the trained network model.

[0049] Within the range of 1 to the preset number of iterations, take integer values ​​for X in sequence and perform the following steps until X equals the preset number of iterations. Then, determine the optimal individual based on the current population and update the update model weights and the update model threshold based on the optimal individual to obtain the optimized network model.

[0050] The historical state information is input into the trained network model to obtain the Xth prediction control time information;

[0051] The fitness of the Xth individual is calculated based on the predicted control time information of the Xth time and the historical control time information corresponding to the historical state information.

[0052] The current population is determined based on the fitness of the Xth individual.

[0053] In another possible implementation, the prediction module is specifically used for:

[0054] A new population is formed by selecting chromosomes with high fitness from X individuals according to a preset algorithm; the preset algorithm is any one of roulette wheel selection, tournament selection, and competitive selection.

[0055] The chromosomes in the new population are subjected to crossover and mutation operations to obtain the current population.

[0056] In another possible implementation, the prediction module is also used for:

[0057] The optimized network model is tested based on the test samples to obtain test results;

[0058] If the test results indicate that the optimized network model does not meet the preset accuracy requirements, then the steps for obtaining the trained network model are repeated.

[0059] In another possible implementation, the prediction module is specifically used for:

[0060] Acquire historical state information and historical control time information of the transmission during upshifting under different operating conditions within a preset time period, and use the historical state information and the historical control time information as sample data;

[0061] The sample data is preprocessed to obtain the training data. The preprocessing includes filtering ideal sample data from the sample data and normalizing the ideal sample data. The historical control time information of the ideal sample data is within a preset reasonable range.

[0062] The training data is divided into training samples and test samples according to a preset ratio. The training samples are used to train the original network model, and the test samples are used to test the optimized network model.

[0063] Thirdly, this application provides a vehicle that includes a data acquisition module, a transmission controller, and an intermediate shaft brake.

[0064] The data acquisition module is used to obtain current status information and send it to the transmission controller. The current status information includes input shaft speed, output shaft speed, transmission housing temperature, current gear, target gear, and intermediate shaft braking slope.

[0065] The transmission controller is configured to receive the current status information, predict control time information based on the current status information, and generate control commands for sending to the intermediate shaft brake based on the control time information, wherein the control time information includes braking start time and total braking time; the transmission controller is also configured to execute the method as described in any of the first aspects;

[0066] The intermediate shaft brake is used to receive the control command and respond to the control command, start braking at the braking start time, and end braking when the total braking time is reached.

[0067] Fourthly, this application provides a transmission controller, including a processor and a memory communicatively connected to the processor;

[0068] The memory stores computer-executed instructions;

[0069] The processor executes computer execution instructions stored in the memory to implement the method as described in any of the first aspects.

[0070] Fifthly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the switching frequency adjustment method as described in any of the first aspects above.

[0071] This application provides a vehicle transmission braking control method, device, and vehicle. The method is executed by the vehicle transmission controller, which acquires current state information in real time and inputs it into an optimized network model to obtain control time information. The transmission controller generates control commands for controlling the intermediate axle brake based on the control time information. Specifically, the control time information includes the braking start time and the total braking time. The control commands instruct the intermediate axle brake to begin braking at the braking start time and to stop braking when the total braking time is reached.

[0072] The method described in this application allows the vehicle's transmission to predict the braking start time and total braking time by using real-time input shaft speed, output shaft speed, transmission housing temperature, current gear, target gear, and intermediate shaft brake slope as references during upshifting. This avoids errors in braking time caused by decreased braking effectiveness due to wear, and improves the control accuracy of the intermediate shaft brake. Simultaneously, it helps to ensure that the speed difference between the sliding sleeve and the gears on the two shafts of the target gear is within a reasonable range, resulting in a smooth, shock-free, and rapid shifting process. Attached Figure Description

[0073] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0074] Figure 1 A schematic diagram of a vehicle frame provided for an embodiment of this application;

[0075] Figure 2 A schematic flowchart of a vehicle transmission braking control method provided in an embodiment of this application;

[0076] Figure 3 A flowchart illustrating a training method for a primitive network model provided in this application embodiment;

[0077] Figure 4 A flowchart illustrating an optimization method for a trained network model provided in an embodiment of this application;

[0078] Figure 5 This is a schematic diagram of the structure of a vehicle transmission braking control device provided in an embodiment of this application;

[0079] Figure 6 An electronic device provided in an embodiment of this application.

[0080] The accompanying drawings have illustrated specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to specific embodiments. Detailed Implementation

[0081] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0082] When an automatic transmission shifts gears, the different gear ratios result in a significant speed difference between the engagement sleeve and the gear to be engaged, making shifting difficult. To ensure a fast and smooth shifting process, an auxiliary device is needed to keep the speed difference within a reasonable range.

[0083] Currently, commercial vehicles use sliding sleeve transmissions instead of synchronized transmissions to reduce transmission costs and improve shifting efficiency. However, sliding sleeve transmissions require an auxiliary device to adjust the speed difference between the engaging sleeve and the target gear; otherwise, an unreasonable speed difference can lead to decreased shifting quality or even shifting failure. Because commercial vehicles have their own high-pressure air source, compressed gas is used as the power source for the auxiliary device, and the controller controls the operation of the auxiliary device by controlling the compressed gas.

[0084] The current control strategy of the intermediate shaft brake assist device in automatic transmissions is based on the speed difference between the engagement sleeve and the gear to be engaged during the shifting process, which controls the action of the intermediate shaft brake. During gear shifting while driving, the transmission controller issues a control command to open the solenoid valve of the intermediate shaft brake's air inlet. Compressed gas enters the working cylinder inside the brake, and the resulting pressure overcomes the resistance of the return spring, rapidly squeezing the friction pads. The friction pads interact with the intermediate shaft, causing the intermediate shaft speed to decrease. When the detected speed difference reaches a certain value, the transmission controller issues a control signal to close the air inlet solenoid valve and open the exhaust solenoid valve. The friction pads, along with the return spring, discharge the high-pressure gas in the cylinder to the atmosphere. The intermediate shaft brake and the intermediate shaft gradually lose their force and no longer generate friction, thus completing the intermediate shaft braking.

[0085] In known technologies, most intermediate shaft brakes use fixed parameters to control their operation. However, factors such as wear of the friction pads during use, oil churning resistance affected by oil temperature, frictional resistance of the seals, and the speed difference between the engaging sleeve and the gear to be engaged can all affect the control accuracy of the intermediate shaft brake. If fixed parameters are used to control the intermediate shaft brake, the braking accuracy will be poor, thus affecting the shifting quality. Therefore, there is an urgent need to design a new control method for intermediate shaft brakes to improve the accuracy of their operating time.

[0086] This application provides a vehicle transmission braking control method, device, and vehicle to solve the aforementioned problems. The method is executed by the vehicle's transmission controller. The transmission controller trains and optimizes the original network model using historical state information and historical control time information during upshifts under different operating conditions within a preset time period to obtain an optimized network model. Based on this optimized network model, the transmission controller obtains accurate control time information by real-time acquisition of input shaft speed, output shaft speed, transmission housing temperature, current gear, target gear, and intermediate shaft brake slope, and inputs these parameters into the optimized network model. The transmission controller then generates braking commands based on this control time information to instruct the intermediate shaft brake to perform braking control according to the commands.

[0087] The method described in this application allows the intermediate shaft brake to adaptively adjust to follow the real-time status of the transmission, thereby improving the control precision of the intermediate shaft brake and enhancing shift quality.

[0088] The following detailed description of some embodiments of this application is provided in conjunction with the accompanying drawings. Where the embodiments do not conflict, the following embodiments and features thereof can be combined with each other.

[0089] Figure 1 A schematic diagram of a vehicle frame provided in an embodiment of this application, such as... Figure 1 As shown, the vehicle in this application includes a data acquisition module 101, a transmission controller 102, and an intermediate shaft brake 103. The data acquisition module 101 and the transmission controller 102 are communicatively connected, and the transmission controller 102 and the intermediate shaft brake 103 are communicatively connected.

[0090] Specifically, the data acquisition module 101 is used to collect current status information and send it to the transmission controller 102. The current status information includes the input shaft speed, output shaft speed, transmission housing temperature, current gear of the transmission, target gear of the transmission, and intermediate shaft braking slope.

[0091] In this embodiment, the vehicle further includes an input shaft speed sensor, an output shaft speed sensor, an oil temperature sensor, and a memory. The memory stores the slope of the intermediate shaft brake after the previous braking action. The input terminals of the data acquisition module 101 are connected to the output terminals of the input shaft speed sensor, the output shaft speed sensor, the oil temperature sensor, and the memory, respectively, to acquire the input shaft speed, the output shaft speed, the transmission housing temperature, and the intermediate shaft brake slope. The input terminal of the data acquisition module 101 is also connected to the output terminal of the transmission to acquire the current gear and the target gear of the transmission.

[0092] The gearbox controller 102 receives current status information, predicts control time information based on the current status information, and generates control commands to be sent to the intermediate shaft brake based on the control time information. The control time information includes the braking start time and the total braking time. The intermediate shaft brake 103 receives the control commands and responds to them, initiating braking at the braking start time and ending braking when the total braking time is reached.

[0093] In the vehicle described in this application, the data acquisition module 101 collects current status information in real time and sends it to the transmission controller 102. Upon receiving the current status information, the transmission controller 102 predicts control time information based on the current status information and generates a control command based on the obtained control time information, which is then sent to the intermediate axle brake 103. Upon receiving the control command, the intermediate axle brake 103 responds by initiating braking at the braking start time and ending braking when the total braking time is reached. Therefore, in the vehicle described in this application, the intermediate axle brake 103 adaptively adjusts the braking time based on the real-time status of the transmission during braking control, thereby facilitating precise braking control and improving shift quality.

[0094] Figure 2 This is a flowchart illustrating a vehicle transmission braking control method provided in an embodiment of this application. Specifically, the method provided in this embodiment is executed by the vehicle's transmission controller to obtain control timing information and generate control commands based on the control timing information. Figure 2 As shown, the method provided in this embodiment includes:

[0095] S201, obtain current status information in real time.

[0096] Specifically, the current status information includes the input shaft speed, output shaft speed, current gear of the transmission, target gear of the transmission, and the slope of the intermediate shaft brake.

[0097] It is understood that the current status information is obtained by the vehicle's data acquisition module. In this embodiment, the transmission controller obtains the current status information collected by the data acquisition module in real time. In practical applications, to reduce the burden on the transmission controller, the current status information can be obtained at preset intervals. The preset interval should not be too long; for example, it can be 30 seconds.

[0098] S202, input the current state information into the optimized network model to obtain control time information.

[0099] The control time information includes braking start time and total braking time. The optimized network model is obtained by training and optimizing the original network model. During training, the original network model takes historical state information as input and historical control time information as output.

[0100] In this embodiment, a backpropagation (BP) neural network is used as the original network model; specifically, a 6-input, 2-output BP neural network is employed. It is understood that the original network model can be trained and optimized by the transmission controller before use, or it can be trained and optimized by any electronic device before being used by the transmission controller. This embodiment does not limit this, as long as the optimized network model can obtain control timing information based on the current state information during use.

[0101] In this embodiment, the braking start time is used to indicate a specific moment, and the total braking time is of any length. For example, the braking start time is 19:00:00, and the total braking time is 30 seconds.

[0102] S203 generates control commands based on the braking start time and total braking time, and sends the control commands to the intermediate shaft brake.

[0103] The control command is used to instruct the intermediate shaft brake to start braking at the braking start time and to stop braking when the total braking time is reached.

[0104] In this embodiment, when the transmission controller obtains the control time information through the optimized network model, it generates an instruction for the intermediate shaft brake to start braking at the braking start time and to stop braking when the total braking time is reached.

[0105] For example, if the transmission controller inputs the current state information into the optimized network model, and the received control time information indicates that the braking start time is 19:00:00 and the total braking time is 30s, then the transmission controller will generate a function to instruct the intermediate shaft brake to start braking at 19:00:00 and end braking at 19:00:30.

[0106] In this embodiment, the transmission controller acquires current state information in real time, predicts control time information based on the current state information and the optimized network model, and generates control commands based on the control time information to precisely instruct the intermediate shaft brake to begin braking at the braking start time and end braking when the total braking time is reached. It can be understood that this embodiment's method performs intermediate shaft braking control based on the real-time state of the transmission to achieve higher braking accuracy, thereby improving shift quality.

[0107] It is understandable that the current state information may also include other factors that may affect the braking of the intermediate shaft brake. As long as the original network model uses the current state information as input and the response control time information as the target output during training and optimization, this embodiment does not limit the number of factors included in the current state information.

[0108] Figure 3 This is a flowchart illustrating a training method for an original network model provided in this application embodiment. Specifically, this embodiment provides a detailed explanation of how to obtain the original network model and how to train it. Figure 3 As shown, the method provided in this embodiment includes:

[0109] S301, acquire the historical status information and historical control time information of the transmission during upshifts under different operating conditions within a preset time period, and use the historical status information and historical control time information as sample data.

[0110] In this embodiment, the historical status information during upshifts under different operating conditions is obtained by the data acquisition module within a preset time period. The transmission controller records and stores the control time information corresponding to each status information as historical control time information. Therefore, the transmission controller obtains historical status information through the data acquisition module and obtains the historical control time information corresponding to the historical status information through its own storage module.

[0111] In practical applications, the data acquisition module can also include a timer. For each historical state, the timer records the historical braking start time when the intermediate shaft brake begins braking and the historical braking total time when the intermediate shaft brake ends braking. Therefore, the transmission controller can obtain historical control time information through the data acquisition module.

[0112] In this embodiment, the preset time period is 2 months. In practical applications, the preset time period can also be 1 month, but this embodiment does not limit it.

[0113] S302, preprocess the sample data to obtain training data. The preprocessing includes screening the ideal sample data from the sample data and normalizing the ideal sample data.

[0114] Among them, the historical control time information of the ideal sample data is within a preset reasonable range.

[0115] In this embodiment, the preset reasonable range is summarized by the user in advance based on the status information. The historical control time within the preset reasonable range can achieve a better braking effect.

[0116] In this embodiment, the ideal sample data is normalized using the following formula, where x min x is the minimum value in the ideal sample data. max x is the maximum value in the ideal sample data. ’ This represents the value of x after normalization.

[0117]

[0118] S303 divides the training data into training samples and test samples according to a preset ratio.

[0119] The training samples are used to train the original network model, and the test samples are used to test the optimized network model.

[0120] In this embodiment, the preset ratio is 3:1, meaning half of the training data is used as training samples, and the other half is used as testing samples. In actual use, the preset ratio can also be 2:1, but this embodiment does not limit it to this. Dividing the training data into training samples and test samples, used to train the network model and test the network model respectively, is beneficial for obtaining a high-precision network model, thereby facilitating the precise control of the intermediate shaft brake.

[0121] S304. Based on the training data, obtain the topology of the original network model. Based on the topology, determine the number of model weights and model thresholds, and initialize the model weights and model thresholds to obtain the initial weights and initial thresholds, thus obtaining the original network model.

[0122] The topology includes an input layer with 6 neurons, a hidden layer with m neurons, and an output layer with 2 neurons.

[0123] Specifically, based on the training data, the network model should have 6 inputs and 2 outputs. Therefore, the original network model's topology should include an input layer with 6 neurons and an output layer with 2 neurons. In this embodiment, the number of neurons in the hidden layer is set to m, and the size of m is adaptively adjusted according to the training results each time.

[0124] Furthermore, based on the number of neurons in the input layer, hidden layer, and output layer, the number of model weights and model thresholds are determined, and the model weights and model thresholds are initialized to obtain the initial weights and initial thresholds, thereby obtaining the original network model.

[0125] S305, input the historical state information from the training samples into the original network model to obtain the Y-th prediction control time information.

[0126] S306, Calculate the Y-th error based on the Y-th predictive control time information and historical control time information.

[0127] S307. Based on the Y-th error, determine the updated model weights and the updated model threshold, and update the original network model based on the updated model weights and the updated model threshold.

[0128] S308, determine whether Y is equal to the preset number of iterations, and / or whether the error of the Yth iteration is within the preset range. If yes, then execute S309; ​​otherwise, execute S305.

[0129] S309, use the current network model as the trained network model.

[0130] In this embodiment, for each training sample, its historical state information is input into the original network model to obtain the Y-th prediction control time information. The transmission controller calculates the Y-th error using the following formula and performs error backpropagation based on a gradient descent strategy to obtain updated model weights and updated model thresholds, thereby updating the original network model.

[0131]

[0132] In the above formula, E is the Y-th error, y out Let Y be the predicted control time information for the Yth time, and y be the historical control time information corresponding to the historical state information.

[0133] In this embodiment, the transmission controller obtains the updated model weights and updated model thresholds using weight adjustment formulas and threshold adjustment formulas. Specifically, the weight adjustment formulas and threshold adjustment formulas are as follows:

[0134]

[0135]

[0136] In the above formula, Δω is used to represent the updated model weights, ω is used to represent the initial weights, η is the neural network learning rate, Δb is used to represent the updated model threshold, and b is used to represent the initial threshold.

[0137] In this embodiment, the transmission controller first determines the topology of the original network model, as well as the number of model weights and thresholds, based on the number of input and output parameters. The original network model is then obtained by initializing the model weights and thresholds. Next, the transmission controller trains the original network model using training samples from the training data. The error is calculated by using the predicted control time information and the historical control time corresponding to the historical state information obtained during training. Finally, the model weights and thresholds are continuously adjusted based on the error until the error meets the termination training condition and / or reaches a preset number of iterations, resulting in a trained network model.

[0138] Figure 4 This document provides a flowchart of a method for optimizing a trained network model, based on an embodiment of this application. Specifically, this embodiment details how to optimize a trained network model, building upon the preceding embodiments. Figure 4 As shown, the method in this embodiment includes:

[0139] S401: Based on the trained network model, a preset number of chromosomes are randomly generated to form a population.

[0140] In this context, chromosomes serve as the updated model weights and thresholds for the trained network model.

[0141] S402, input the historical state information included in the training samples into the trained network model to obtain the Xth prediction control time information.

[0142] S403, calculate the fitness of the Xth individual based on the historical control time information corresponding to the Xth predicted control time information and the historical state information.

[0143] S404, according to the preset algorithm, select some chromosomes with high fitness from the fitness of X individuals to form a new population.

[0144] The preset algorithm is any one of the following: roulette wheel selection method, tournament selection method, and competitive selection method.

[0145] S405 involves performing crossover and mutation operations on the chromosomes in the new population to obtain the current population.

[0146] S406, determine if X is equal to the preset number of iterations. If yes, execute S407; otherwise, execute S402.

[0147] S407: Determine the optimal individual based on the current population, and update the model weights and thresholds based on the optimal individual to obtain a trained network model.

[0148] S408, Test the optimized network model based on the test samples to obtain the test results; if the test results indicate that the optimized network model does not meet the preset accuracy requirements, repeat the steps to obtain the trained network model.

[0149] In this embodiment, the trained network model is optimized using a genetic algorithm to obtain an optimized network model. Specifically, the transmission controller determines the number of parameters to be optimized based on the trained network model, such that a set of parameters constitutes a chromosome, and a preset number of chromosomes are randomly generated to form a population.

[0150] Furthermore, the trained network model is trained using training samples, and the individual fitness is calculated based on the historical time control information corresponding to the Xth prediction control time information and the input historical state information. Specifically, in this embodiment, it is calculated using the following formula:

[0151]

[0152] In the above formula, y outi For the Xth prediction time control information corresponding to the i-th group, y i Let k be the historical control time information included in the i-th group of sample data, and k be the number of sample data.

[0153] In this embodiment, a roulette wheel selection algorithm is used to select individuals with high fitness to form a new population. Specifically, the probability of each individual j being selected is: Where n is the number of individuals in the population, g j Let be the fitness of individual j.

[0154] Furthermore, two chromosomes are randomly selected from the population, and the crossover probability is used to determine whether to perform crossover. The steps for real crossover are as follows:

[0155]

[0156] In the above formula, a mj a is the real number at position j on the m-th chromosome. nj It is a real number at position j on the nth chromosome, and rand is a random number between (0,1).

[0157] In this embodiment, an individual is randomly selected from the population, and the mutation process is determined based on the mutation probability.

[0158]

[0159]

[0160] In the above formula, aij It is the real number at position j of the i-th chromosome, and g is the current iteration number. max Let a be the maximum number of iterations. max It is a ij The upper limit of the value of a min It is a ij The lower bound of the value, rand and rand ’ It is a random number between (0,1).

[0161] In this embodiment, the current iteration number is compared with the preset iteration number to determine whether to end the iteration. If the iteration ends, the best individual in the current population obtained in the last iteration is taken as the optimal solution, resulting in an optimized network model. Otherwise, the iteration continues.

[0162] In this embodiment, after obtaining the optimized network model, it is tested using test samples to obtain test results. It is understood that if the test results indicate that the error is within a preset range, it means that the optimized network model has achieved the training objective. Otherwise, it means that the optimized network model has not achieved the training objective. In this embodiment, for network models that have not achieved the training objective, the training, optimization, and testing steps are repeated until the test results indicate that the error of the optimized network model is within the preset range.

[0163] The method in this embodiment further optimizes the trained network model to obtain an optimized network model, thereby effectively improving the model accuracy. Furthermore, the optimized network model is tested using test samples to verify its accuracy. If the test results are unsatisfactory, the training, optimization, and testing steps are repeated until the accuracy of the optimized network model meets the preset requirements.

[0164] The above embodiments describe a vehicle transmission braking control method from the perspective of process flow. The following embodiments describe a vehicle transmission braking control device from the perspective of virtual module or virtual unit. For details, please refer to the following embodiments.

[0165] This application provides a vehicle transmission braking control device, such as... Figure 5 As shown, the device includes an acquisition module 51, a prediction module 52, and a control module 53, wherein,

[0166] The acquisition module 51 is used to acquire the current status information in real time. The current status information includes the input shaft speed, output shaft speed, gearbox housing temperature, current gear of the gearbox, target gear of the gearbox, and intermediate shaft brake slope.

[0167] The prediction module 52 is used to input the current state information into the optimized network model to obtain control time information, which includes braking start time and total braking time. The optimized network model is obtained by training and optimizing the original network model. During training, the original network model takes historical state information as input and historical control time information as the target output.

[0168] The control module 53 is used to generate control commands based on the braking start time and the total braking time, and send the control commands to the intermediate shaft brake. The control commands are used to instruct the intermediate shaft brake to start braking at the braking start time and to stop braking when the total braking time is reached.

[0169] In another possible implementation of this application embodiment, the prediction module 52 is further used for:

[0170] Acquire training data, which includes training samples. The training samples include historical state information and historical control time information corresponding to the historical state information.

[0171] Take integer values ​​for Y within the range of 1 to the preset number of iterations, and perform the following steps until the error of the Yth iteration satisfies the termination training condition, and / or Y equals the preset number of iterations, to obtain the trained network model; wherein, the termination training condition is: the error of the Yth iteration is within the preset range;

[0172] The historical state information is input into the original network model to obtain the Y-th prediction control time information;

[0173] Calculate the Y-th error based on the Y-th predicted control time information and the historical control time information;

[0174] Based on the Y-th error, determine the updated model weights and the updated model threshold, and update the original network model based on the updated model weights and the updated model threshold.

[0175] In another possible implementation of this embodiment, the prediction module 52 is further configured to:

[0176] Based on the training data, the topology of the original network model is obtained; the topology includes an input layer with 6 neurons, a hidden layer with m neurons, and an output layer with 2 neurons.

[0177] Based on the topology, determine the number of model weights and model thresholds, and initialize the model weights and model thresholds to obtain the initial weights and initial thresholds, thus obtaining the original network model.

[0178] In another possible implementation of this application embodiment, the prediction module 52 is further used for:

[0179] Based on the trained network model, a predetermined number of chromosomes are randomly generated to form a population; the chromosomes are used as the updated model weights and updated model thresholds of the trained network model.

[0180] Within the range of 1 to the preset number of iterations, take integer values ​​for X in sequence and perform the following steps until X equals the preset number of iterations. Then, determine the optimal individual based on the current population and update the model weights and thresholds based on the optimal individual to obtain the optimized network model.

[0181] The historical state information is input into the trained network model to obtain the Xth prediction control time information;

[0182] The fitness of the Xth individual is calculated based on the historical control time information corresponding to the Xth prediction control time information and the historical state information.

[0183] The current population is determined based on the fitness of the Xth individual.

[0184] In another possible implementation of this application embodiment, the prediction module 52 is specifically used for:

[0185] Based on the trained network model, a predetermined number of chromosomes are randomly generated to form a population; the chromosomes are used as the updated model weights and updated model thresholds of the trained network model.

[0186] Within the range of 1 to the preset number of iterations, take integer values ​​for X in sequence and perform the following steps until X equals the preset number of iterations. Then, determine the optimal individual based on the current population and update the model weights and thresholds based on the optimal individual to obtain the optimized network model.

[0187] The historical state information is input into the trained network model to obtain the Xth prediction control time information;

[0188] The fitness of the Xth individual is calculated based on the historical control time information corresponding to the Xth prediction control time information and the historical state information.

[0189] The current population is determined based on the fitness of the Xth individual.

[0190] In another possible implementation of this application embodiment, the prediction module 52 is further used for:

[0191] The optimized network model is tested based on the test samples to obtain the test results;

[0192] If the test results indicate that the optimized network model does not meet the preset accuracy requirements, then repeat the steps to obtain the trained network model.

[0193] In another possible implementation of this application embodiment, the prediction module 52 is specifically used for:

[0194] Acquire historical status information and historical control time information of the transmission during upshifting under different operating conditions within a preset time period, and use the historical status information and historical control time information as sample data.

[0195] The sample data is preprocessed to obtain training data. The preprocessing includes selecting ideal sample data from the sample data and normalizing the ideal sample data. The historical control time information of the ideal sample data is within a preset reasonable range.

[0196] The training data is divided into training samples and test samples according to a preset ratio. The training samples are used to train the original network model, and the test samples are used to test the optimized network model.

[0197] The vehicle transmission braking control device provided in this application embodiment is applicable to the above method embodiment, and will not be described again here.

[0198] This application provides a transmission controller, such as... Figure 6 As shown, Figure 6 The transmission controller shown includes a processor 61 and a memory 62. The processor 61 and the memory 62 are connected, for example, via a bus 63. Optionally, the transmission controller may also include a transceiver 64. It should be noted that in practical applications, the transceiver 64 is not limited to one, and the structure of this transmission controller does not constitute a limitation on the embodiments of this application.

[0199] Processor 61 may be a central processing unit (CPU), a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 61 may also be a combination that implements computational functions, such as a combination of one or more microprocessors 61, a combination of a DSP and a microprocessor 61, etc.

[0200] Bus 63 may include a pathway for transmitting information between the aforementioned components. Bus 63 may be a Peripheral Component Interconnect (PCI) bus 63 or an Extended Industry Standard Architecture (EISA) bus 63, etc. Bus 63 can be divided into address bus 63, data bus 63, control bus 63, etc. For ease of representation, Figure 6 The bus 63 is represented by only one thick line, but this does not mean that there is only one bus 63 or one type of bus 63.

[0201] The memory 62 may be a read-only memory 62 (ROM) or other type of static storage device capable of storing static information and instructions, a random access memory 62 (RAM) or other type of dynamic storage device capable of storing information and instructions, or an electrically erasable programmable read-only memory 62 (EEPROM), a compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), a magnetic disk storage medium or other magnetic storage device, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto.

[0202] The memory 62 is used to store application code that executes the solution of this application, and its execution is controlled by the processor 61. The processor 61 is used to execute the application code stored in the memory 62 to implement the content shown in the foregoing method embodiments.

[0203] This application also provides a computer-readable storage medium, which may include various media capable of storing program code, such as a USB flash drive, a portable hard drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk. Specifically, the computer-readable storage medium stores program instructions, which are used to implement the methods in the above embodiments.

[0204] This application also provides a computer program product, including a computer program that, when executed by a processor, implements the technical solution of the above method embodiments. Its implementation principle and technical effects are similar, and will not be repeated here.

[0205] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the claims.

[0206] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A vehicle transmission braking control method, characterized in that, Performed by the transmission brake, the method includes: Real-time acquisition of current status information, including input shaft speed, output shaft speed, gearbox housing temperature, current gear of the gearbox, target gear of the gearbox, and intermediate shaft brake slope; The current state information is input into the optimized network model to obtain control time information, which includes braking start time and total braking time. The optimized network model is obtained by training and optimizing the original network model. During training, the original network model takes historical state information as input and historical control time information as the target output. Based on the braking start time and the total braking time, a control command is generated and sent to the intermediate shaft brake. The control command is used to instruct the intermediate shaft brake to start braking at the braking start time and to stop braking when the total braking time is reached. Before inputting the current state information into the optimized network model, the method further includes: Acquire training data, which includes training samples, and the training samples include historical state information and historical control time information corresponding to the historical state information; Take integer values ​​for Y within the range of 1 to a preset number of iterations, and perform the following steps until the error of the Yth iteration satisfies the termination training condition, and / or Y is equal to the preset number of iterations, to obtain the trained network model; wherein, the termination training condition is: the error of the Yth iteration is within the preset range; The historical state information is input into the original network model to obtain the Y-th prediction control time information; The Y-th error is calculated based on the Y-th predicted control time information and the historical control time information. Based on the Y-th error, determine the updated model weights and the updated model threshold, and update the original network model based on the updated model weights and the updated model threshold.

2. The method according to claim 1, characterized in that, Before inputting the historical control time information into the original network model, the method further includes: Based on the training data, the topology of the original network model is obtained; the topology includes an input layer with 6 neurons, a hidden layer with m neurons, and an output layer with 2 neurons. Based on the topology, the number of model weights and model thresholds is determined, and the model weights and model thresholds are initialized to obtain initial weights and initial thresholds, thereby obtaining the original network model.

3. The method according to claim 1, characterized in that, Before inputting the current state information into the optimized network model, the method further includes: Based on the trained network model, a preset number of chromosomes are randomly generated to form a population; the chromosomes serve as the updated model weights and the updated model thresholds of the trained network model. Within the range of 1 to the preset number of iterations, take integer values ​​for X in sequence and perform the following steps until X equals the preset number of iterations. Then, determine the optimal individual based on the current population and update the update model weights and the update model threshold based on the optimal individual to obtain the optimized network model. The historical state information is input into the trained network model to obtain the Xth prediction control time information; The fitness of the Xth individual is calculated based on the predicted control time information of the Xth time and the historical control time information corresponding to the historical state information. The current population is determined based on the fitness of the Xth individual.

4. The method according to claim 3, characterized in that, The step of determining the current population based on the fitness of the Xth individual includes: A new population is formed by selecting chromosomes with high fitness from X individuals according to a preset algorithm; the preset algorithm is any one of roulette wheel selection, tournament selection, and competitive selection. The chromosomes in the new population are subjected to crossover and mutation operations to obtain the current population.

5. The method according to claim 1, characterized in that, The training data also includes test samples, and after obtaining the optimized network model, the method further includes: The optimized network model is tested based on the test samples to obtain test results; If the test results indicate that the optimized network model does not meet the preset accuracy requirements, then the steps for obtaining the trained network model are repeated.

6. The method according to claim 1, characterized in that, The acquisition of training data includes: Acquire historical state information and historical control time information of the transmission during upshifting under different operating conditions within a preset time period, and use the historical state information and the historical control time information as sample data; The sample data is preprocessed to obtain the training data. The preprocessing includes filtering ideal sample data from the sample data and normalizing the ideal sample data. The historical control time information of the ideal sample data is within a preset reasonable range. The training data is divided into training samples and test samples according to a preset ratio. The training samples are used to train the original network model, and the test samples are used to test the optimized network model.

7. A vehicle transmission braking control device, characterized in that, include: The acquisition module is used to acquire current status information in real time, including input shaft speed, output shaft speed, gearbox housing temperature, current gearbox gear, target gearbox gear, and intermediate shaft brake slope. The prediction module is used to input the current state information into the optimized network model to obtain control time information, which includes braking start time and total braking time. The optimized network model is obtained by training and optimizing the original network model. During training, the original network model takes historical state information as input and historical control time information as the target output. The control module is configured to generate a control command based on the braking start time and the total braking time, and send the control command to the intermediate shaft brake. The control command is configured to instruct the intermediate shaft brake to start braking at the braking start time and to stop braking when the total braking time is reached. The prediction module is also used for: Acquire training data, which includes training samples, and the training samples include historical state information and historical control time information corresponding to the historical state information; Take integer values ​​for Y within the range of 1 to a preset number of iterations, and perform the following steps until the error of the Yth iteration satisfies the termination training condition, and / or Y is equal to the preset number of iterations, to obtain the trained network model; wherein, the termination training condition is: the error of the Yth iteration is within the preset range; The historical state information is input into the original network model to obtain the Y-th prediction control time information; The Y-th error is calculated based on the Y-th predicted control time information and the historical control time information. Based on the Y-th error, determine the updated model weights and the updated model threshold, and update the original network model based on the updated model weights and the updated model threshold.

8. A vehicle, characterized in that, The vehicle includes a data acquisition module, a transmission controller, and an intermediate shaft brake; The data acquisition module is used to collect current status information and send it to the transmission controller; the current status information includes input shaft speed, output shaft speed, transmission housing temperature, current gear of the transmission, target gear of the transmission, and intermediate shaft braking slope; The transmission controller is configured to receive the current status information, predict control time information based on the current status information, and generate control commands for sending to the intermediate shaft brake based on the control time information, wherein the control time information includes braking start time and total braking time; the transmission controller is also configured to perform the method as described in any one of claims 1-6; The intermediate shaft brake is used to receive the control command and respond to the control command, start braking at the braking start time, and end braking when the total braking time is reached.

9. A gearbox controller, characterized in that, Includes a processor and a memory communicatively connected to the processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory to implement the method as described in any one of claims 1 to 6.

Images