Automatic parking function control method and system and electronic equipment

Abstract

The invention belongs to the field of vehicles, and discloses an automatic parking function control method and system and electronic equipment, and the method comprises the steps: obtaining vehicle operation information; inputting the vehicle operation information into a preset anomaly prediction model to obtain a prediction result output by the anomaly prediction model; and outputting a corresponding result according to a prediction result. The road condition information and the vehicle control information are acquired in real time, and the preset abnormal prediction model is utilized to predict the abnormal activation or non-activation condition possibly occurring in the automatic parking function, so that the accurate control instruction is output to control the automatic parking function to be turned on or turned off in advance; according to the method, the safety and comfort of automatic parking are remarkably improved, the risk of misoperation is reduced, the driving convenience is improved, the driving experience is optimized, the intelligence and automation of the automatic parking function are achieved, the method adapts to variable traffic environments, and safer, more comfortable and more convenient driving experience is brought to a driver.

Description

Technical Field

[0001] The present invention relates to the field of vehicles, and in particular to a method, system, electronic device and computer-readable storage medium for controlling an automatic parking function. Background Art

[0002] With the rapid development of automotive electronic technology, automatic parking system has become an important function in modern automobiles. It automatically controls the braking force of the vehicle through an electronic control unit to improve driving comfort and safety.

[0003] The automatic parking function can maintain braking force for a certain period of time after the driver stops the vehicle by using the brake pedal, thus eliminating the need for the driver to continuously step on the brake pedal. At present, most automatic parking function control systems rely on fixed control logic and parameter settings, which cannot be automatically adjusted in real time according to changes in road conditions and user driving habits. For example, in congested urban road conditions, the automatic parking function may be mistakenly activated due to frequent starts and stops, causing driving frustration and increasing the driver's operating burden. On slopes or slippery roads, the system may not be able to accurately identify road conditions, resulting in abnormal inactivation of the automatic parking function, thereby increasing the risk of slipping. This not only affects the driving experience, but may also cause safety issues.

[0004] Therefore, how to reduce the abnormal probability of the automatic parking function is a technical problem to be solved by those skilled in the art. Summary of the invention

[0005] In order to solve the problem of abnormal activation or abnormal inactivation of the automatic parking function in the prior art, the present invention provides a method, system, electronic device and computer-readable storage medium for controlling the automatic parking function.

[0006] A method for controlling an automatic parking function, comprising: Acquiring vehicle operation information; the vehicle operation information includes road condition information and vehicle control information; Inputting the vehicle operation information into a preset abnormality prediction model to obtain a prediction result output by the abnormality prediction model; the prediction result includes a probability of function abnormality activation and a probability of function abnormality non-activation; A corresponding control instruction is output according to the prediction result; when the control instruction is executed, the automatic parking function is controlled to be turned on or off.

[0007] Optionally, the process of establishing the abnormality prediction model includes: Acquire historical vehicle operation information, and perform abnormality identification on the historical vehicle operation information to obtain abnormality identification results; the abnormality identification results include abnormal function activation and abnormal function inactivation; Performing data cleaning and formatting on the historical vehicle operation information, and constructing a training data set based on the processed historical vehicle operation information; Establishing an initial model, and inputting the training data set into the initial model for forward propagation processing to obtain an output value; A first loss value between the output value and the abnormality recognition result is calculated, and the parameters in the initial model are adjusted according to the first loss value through a back propagation algorithm to obtain the abnormality prediction model.

[0008] Optionally, the performing abnormality identification on the historical vehicle operation information to obtain an abnormality identification result includes: determining status information of an automatic parking function according to the historical vehicle operation information; When the automatic parking function is in an activated state and satisfies a first preset condition, determining that the abnormality identification result is abnormal activation of the function; The first preset condition includes: receiving an input automatic parking function shut-off signal; and / or the duration of the automatic parking function being in an activated state is less than a first threshold; and / or the automatic parking function is repeatedly activated within a set duration; When the automatic parking function is in an inactivated state and meets a second preset condition, determining that the abnormality identification result is that the function is abnormal and inactivated; The second preset condition includes: the automatic parking function directly jumps to the activation state after receiving an input automatic parking function activation signal; and / or, the automatic parking function activation signal is received after the vehicle is powered on and before starting.

[0009] Optionally, the method further includes: Performing abnormality identification on the vehicle operation information to obtain real-time abnormality identification results; A second loss value between the real-time anomaly identification result and the prediction result is calculated, and the deviation prediction model is optimized according to the second loss value.

[0010] Optionally, outputting a corresponding control instruction according to the prediction result includes: When the probability of activation of the abnormal function exceeds a second threshold, outputting an automatic parking function shut-down instruction; When the function abnormality inactivation probability exceeds a third threshold, an automatic parking function activation instruction is output.

[0011] Optionally, after outputting a corresponding control instruction according to the prediction result, the method further includes: When the control instruction is the automatic parking function activation instruction, judging whether the automatic parking function is in the activation state according to the vehicle operation information; If not, turn on the automatic parking function; determining whether the automatic parking function is activated; If yes, then after the automatic parking function exits, turn off the automatic parking function; If not, determining whether the time of receiving the automatic parking function closing instruction exceeds a fourth threshold; If exceeded, the automatic parking function will be turned off.

[0012] Optionally, after outputting a corresponding control instruction according to the prediction result, the method further includes: When the control instruction is the automatic parking function off instruction, judging whether the automatic parking function is in an off state according to the vehicle operation information; If not, turn off the automatic parking function; determining whether the time of receiving the automatic parking function shut-down instruction exceeds a fifth threshold; If it exceeds, the automatic parking function will be turned on.

[0013] A system for controlling an automatic parking function includes a sensor, a chassis controller, an actuator, a satellite positioning module, a network connection module, a vehicle controller and a cloud computing platform, wherein: The sensor, the actuator, the satellite positioning module, the network connection module, the vehicle controller and the cloud computing platform are respectively directly or indirectly connected to the chassis controller for communication; The sensor, the network connection module, the satellite positioning module and the vehicle controller obtain vehicle operation information; The cloud computing platform trains an anomaly prediction model; The actuator executes the control instruction; The chassis controller includes a processor and a memory in communication with the processor, wherein the memory is used to store at least one instruction, and when the instruction is loaded and executed by the processor, the method for controlling the automatic parking function as described in any one of the above is implemented.

[0014] An electronic device, comprising: A processor and a memory, wherein the memory is used to store at least one instruction, and when the instruction is loaded and executed by the processor, the method for controlling the automatic parking function as described in any one of the above is implemented.

[0015] A computer-readable storage medium stores a computer program, wherein the computer program, when executed by a processor, implements the method for controlling an automatic parking function as described in any one of the above.

[0016] The method for controlling the automatic parking function provided by the embodiment of the present invention obtains vehicle operation information; inputs the vehicle operation information into a preset abnormal prediction model to obtain a prediction result output by the abnormal prediction model; outputs a corresponding control instruction according to the prediction result; and controls the automatic parking function to be turned on or off when the control instruction is executed. The present invention obtains road condition information and vehicle control information in real time, and uses a preset abnormal prediction model to predict the abnormal activation or inactivation of the automatic parking function, thereby outputting accurate control instructions to control the opening or closing of the automatic parking function in advance, so as to reduce the probability of abnormal activation or inactivation of the automatic parking function. The safety and comfort of automatic parking are significantly improved, the risk of misoperation is reduced, the driving convenience is improved, and the driving experience is optimized, the intelligence and automation of the automatic parking function are realized, and the automatic parking function is adapted to the changing traffic environment, bringing a safer, more comfortable and convenient driving experience to the driver. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0018] Figure 1 A flow chart of a method for controlling an automatic parking function provided by an embodiment of the present invention; Figure 2 A flowchart of a process for establishing an abnormality prediction model provided by an embodiment of the present invention; Figure 3 A flowchart of another method for controlling an automatic parking function provided by an embodiment of the present invention; Figure 4 A flowchart of another method for controlling an automatic parking function provided by an embodiment of the present invention; Figure 5 A schematic diagram of the structure of an automatic parking function control system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0019] In order to better understand the technical solution of the present invention, the embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0020] It should be clear that the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] The terms used in the embodiments of the present invention are only for the purpose of describing specific embodiments, and are not intended to limit the present invention. The singular forms "a", "said" and "the" used in the embodiments of the present invention and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings.

[0022] It should be understood that the term "and / or" used in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship.

[0023] With the rapid development of automotive electronic technology, the automatic parking system has become an important function in modern cars. It automatically controls the braking force of the vehicle through an electronic control unit to improve driving comfort and safety. The automatic parking function can maintain the braking force for a certain period of time after the driver stops the vehicle by using the brake pedal, so that the driver does not need to continuously step on the brake pedal. At present, most automatic parking function control systems rely on fixed control logic and parameter settings, which cannot be automatically adjusted in real time according to changes in road conditions and user driving habits. For example, in congested urban road conditions, the automatic parking function may be mistakenly activated due to frequent start-stop, resulting in driving frustration and increasing the driver's operating burden. On slopes or slippery roads, the system may not be able to accurately identify the road conditions, resulting in abnormal inactivation of the automatic parking function, thereby increasing the risk of slipping. This not only affects the driving experience, but may also cause safety issues.

[0024] In addition, the existing automatic parking control system lacks an intelligent judgment mechanism for function activation and deactivation. In some cases, the driver needs to manually operate to turn on or off the automatic parking function, which reduces the convenience and safety of driving to a certain extent. Therefore, the present invention provides a method for controlling an automatic parking function to solve the above problems.

[0025] Please refer to Figure 1 , is a flow chart of a method for controlling an automatic parking function provided by an embodiment of the present invention, comprising the following steps: Step S01, obtaining vehicle operation information.

[0026] In this embodiment, the vehicle operation information may include road condition information and vehicle control information, wherein: Traffic condition information may include: whether there is a traffic jam, whether there is a red light, satellite positioning location, slope, ambient brightness, rainfall, etc. Vehicle control information may include: vehicle speed, gear position, brake pedal information, accelerator pedal information, brake master cylinder pressure, drive torque, automatic parking function status, etc.

[0027] In this embodiment, the vehicle operation information can provide the necessary data support for turning the automatic parking function on or off. The system can accurately judge the current state of the vehicle and the upcoming road conditions based on the vehicle operation information, so as to make reasonable control decisions. Through the vehicle operation information, the automatic parking system can predict and avoid unexpected activation or non-activation situations that may occur under specific road conditions, ensuring that the vehicle can achieve safe and smooth parking and starting in various driving scenarios, significantly improving driving safety and comfort.

[0028] In some embodiments, after the vehicle operation information is acquired, the vehicle operation information may be cleaned and formatted so that the abnormal prediction model can make deviation predictions based on it.

[0029] Step S02: input the vehicle operation information into a preset abnormality prediction model to obtain a prediction result output by the abnormality prediction model.

[0030] In this embodiment, the real-time monitored vehicle operation information is input as an input variable into a pre-built abnormal prediction model. The abnormal prediction model is based on historical data and machine learning algorithms, and can predict the abnormal activation or inactivation of the automatic parking function according to the input vehicle operation information. The prediction results include the probability of abnormal function activation and the probability of abnormal function inactivation, which are calculated by the model based on the vehicle operation information and the learned pattern.

[0031] This embodiment enables the system to more intelligently adapt to different driving scenarios and road conditions, reduce safety risks caused by system misjudgment, and improve driving comfort and convenience. By predicting the probability of abnormal activation or inactivation of the function, the system can make adjustments in advance and provide more precise control instructions when necessary to ensure the correct execution of the automatic parking function and avoid activating or inactivating the automatic parking function at inappropriate times, thereby reducing the sense of frustration during driving, improving the driving experience, and reducing the risk of rolling.

[0032] Please refer to Figure 2 , is a flow chart of a process of establishing an abnormality prediction model provided by an embodiment of the present invention. In some embodiments, the abnormality prediction model mentioned in step S02, its establishment process may specifically include the following steps: Step S11, obtaining historical vehicle operation information, and performing abnormality identification on the historical vehicle operation information to obtain an abnormality identification result.

[0033] In this embodiment, the historical vehicle operation information includes historical road condition information and historical vehicle control information, and the abnormal identification result includes abnormal function activation and abnormal function inactivation. By analyzing the historical vehicle operation information, this embodiment can identify abnormal activation or inactivation events that may occur in the automatic parking function under specific circumstances. The purpose of abnormal identification is to determine which situations will cause the automatic parking function to perform unpredictably, such as activation when it should not be activated, or failure to activate when it needs to be activated.

[0034] This process provides data support for building a model that can predict abnormal behavior of the automatic parking function. By identifying and marking these abnormal events, a training data set can be created to train the machine learning model so that the model can learn the characteristics of normal and abnormal behavior, so that when encountering similar situations in the future, it can accurately predict whether the automatic parking function will be abnormal, thereby improving the reliability and safety of the automatic parking system.

[0035] In some embodiments, as mentioned in step S11, the abnormality identification of historical vehicle operation information is performed to obtain the abnormality identification result, which may specifically include the following steps: Step S21, determining the status information of the automatic parking function according to the historical vehicle operation information.

[0036] In this embodiment, the historical vehicle operation information includes historical vehicle control information, and the historical vehicle control information includes automatic parking function status information. This embodiment extracts status information related to the automatic parking function from the historical vehicle operation information. Its purpose is to use the status information to identify and record the status of the automatic parking function at a specific moment in the past, including whether it is activated or disabled, so as to evaluate whether the automatic parking function is abnormally activated or abnormally inactivated, so that the abnormal prediction model trained thereby can understand the performance of the automatic parking function under different driving conditions.

[0037] In this embodiment, the purpose of determining the state information of the automatic parking function is to achieve anomaly detection. By analyzing the state information, it is possible to detect whether the function is activated or deactivated at an inappropriate time, thereby identifying potential abnormal behavior. At the same time, the state information enriches the training data set and provides key label data for the anomaly prediction model. The model can learn under what circumstances the function should be activated or deactivated, and under what circumstances these activations or deactivations may be abnormal.

[0038] Step S22: when the automatic parking function is in an activated state and satisfies a first preset condition, determining that the abnormality identification result is abnormal activation of the function.

[0039] In this embodiment, the first preset condition includes: receiving an input automatic parking function shutdown signal; and / or, the duration of the automatic parking function in the activated state is less than a first threshold; and / or, the automatic parking function is repeatedly activated within a set duration.

[0040] This embodiment monitors the status information of the automatic parking function, and marks it as abnormal function activation when the automatic parking function is in an activated state and meets the first preset condition. It can accurately identify the situation where the automatic parking function is activated when it should not be activated. These abnormal activation data can be used to further train and optimize the abnormal prediction model, so that the system can better adapt to different driving scenarios and conditions.

[0041] Step S23, when the automatic parking function is in an inactivated state and satisfies a second preset condition, determining that the abnormality identification result is that the function is abnormal and inactivated.

[0042] In this embodiment, the second preset condition includes: the automatic parking function directly jumps to the activated state after receiving the input automatic parking function start signal; and / or, the automatic parking function start signal is received after the vehicle is powered on and before starting.

[0043] This embodiment monitors the status information of the automatic parking function, and marks it as abnormally inactivated when the automatic parking function is in an inactivated state and meets the second preset condition. It can accurately identify the situation where the automatic parking function is not activated when it should be activated. These abnormally inactivated data can also be used to further train and optimize the abnormal prediction model, so that the system can better adapt to different driving scenarios and conditions.

[0044] Based on the above technical solution, this embodiment can accurately identify abnormal behaviors that may occur in the automatic parking function under specific circumstances. This detailed abnormality identification helps to build a more accurate and efficient abnormality prediction model, thereby further improving the reliability and safety of the automatic parking function.

[0045] Step S12, performing data cleaning and formatting processing on the historical vehicle operation information, and constructing a training data set based on the processed historical vehicle operation information.

[0046] In this embodiment, the data cleaning process includes removing or correcting errors, omissions or inconsistencies in the data to ensure the quality and accuracy of the data. The formatting process ensures that all data follow the same standard and format to accelerate the convergence speed of the learning algorithm and improve the generalization ability of the model. After completing these processes, the cleaned and formatted data are organized into a training data set, which will be provided as input to the initial model to train the model to identify the relationship between vehicle operation information and abnormal function activation events and / or abnormal function deactivation events.

[0047] In some embodiments, normalizing the historical vehicle operation information may include normalizing a non-numeric signal to a numeric value, for example: Is there a traffic jam? The traffic jam signal comes from the cloud map information. The traffic jam is marked as 1; the no traffic jam is marked as 0.

[0048] Is it a red light? The red light signal comes from the cloud map information. The red light is marked as 1; the non-red light is marked as 0.

[0049] Satellite positioning position: The satellite positioning position comes from the satellite positioning module carried by the vehicle, using longitude and latitude coordinates. For example, if the position of a vehicle at a certain moment is 116°23′17″ east longitude and 39°54′27″ north latitude, it is standardized to [116.23.17, 39.54.27].

[0050] Slope: Slope The longitudinal acceleration value measured by the vehicle's acceleration sensor and vehicle acceleration obtained by wheel speed sensors Calculation results: in, is the acceleration due to gravity.

[0051] After the slope is calculated, it is converted into a numerical value, such as a slope with an upward slope of 10% is marked as 10, and a slope with a downward slope of 10% is marked as -10.

[0052] Ambient brightness: divided into three levels according to the ambient brightness. Ambient brightness detection value ≤ 50 lux is considered night and marked as 0; 50 lux < ambient brightness detection value ≤ 10000 lux is considered cloudy and marked as 1; 10000 lux < ambient brightness detection value is considered sunny and marked as 2.

[0053] Rainfall: Rainfall information comes from the rain signal used by the vehicle's windshield wipers and is divided into three levels according to the actual rainfall. No rain is marked as 0; light rain is marked as 1; heavy rain is marked as 2.

[0054] Vehicle speed: The vehicle speed information comes from the value measured by the wheel speed sensor. The actual vehicle speed is converted into a numerical value, such as 10km / h is marked as 10.

[0055] Gear position: The gear position information is obtained based on the actual vehicle gear position. The D gear is marked as 1, the R gear is marked as 2, and the N gear is marked as 0.

[0056] Brake pedal: Convert to a numerical value based on the percentage of brake pedal depression, such as 10% is marked as 10.

[0057] Accelerator pedal: Converts to a numerical value based on the percentage of accelerator pedal depression, such as 50% is marked as 50.

[0058] Brake master cylinder pressure: Convert the value measured by the master cylinder pressure sensor into a numerical value, such as 20 bar is marked as 20.

[0059] Driving torque: Converted into a numerical value based on the actual vehicle driving torque value, such as 20Nm is marked as 20.

[0060] Automatic parking function status: function off is marked as 0; function on but not activated is marked as 1; function activated is marked as 2.

[0061] In summary, the input layer data matrix (1×14) that characterizes the actual vehicle operating conditions and usage scenarios can be obtained in real time, such as: [1 1 116.23.17 39.5427 0 1 0 0 1 10 0 20 0 2] represents a traffic jam, a red light, a location of 116°23′17″ East longitude, 39°54′27″ North latitude, a slope of 0%, cloudy weather, a speed of 0 km / h, D gear, a brake pedal depressed by 10%, an accelerator pedal depressed by 0%, a master cylinder pressure of 20 bar, a drive torque of 0, and the automatic parking function is activated.

[0062] Step S13, establish an initial model, and input the training data set into the initial model for forward propagation processing to obtain an output value.

[0063] In this embodiment, the preprocessed training data set is input into the initial model, and the output value of the model is calculated by the forward propagation algorithm. Forward propagation is a direction of information flow in a neural network, which starts from the input layer, passes through the hidden layer, and finally reaches the output layer. In this process, the neurons of each layer will perform weighted summation on the output of the previous layer according to their weights and biases, and generate new results through the activation function, which serve as the input of the next layer. In this way, the model can learn the relationship between the input data and the output results.

[0064] In some embodiments, the output value may specifically be a 2×1 data matrix, which is used to represent the probability of abnormal function activation and the probability of abnormal function deactivation, wherein: The probability of abnormal function activation can be determined based on the abnormality identification result, with abnormal function activation events marked as 100 and no abnormal function activation events marked as 0; The probability of functional abnormality inactivation can be determined based on the abnormality identification result, with the functional abnormality inactivation event marked as 100 and the functional abnormality inactivation event marked as 0; If the output value is a matrix , it means that the probability of abnormal function activation is 100%, and the probability of abnormal function non-activation is 0%.

[0065] In some embodiments, a model that supports classification problems, such as a decision tree, a support vector machine, or a deep neural network model, can be used as the initial model. Here, the initial model is taken as an example of a neural network model, and the process of establishing the initial model can be: 1) The input layer is the input layer data matrix (1×14); 2) The output layer is the output layer data matrix (2×1); 3) There are many activation functions that can be used, such as sigmoid function, ReLU function, Leaky ReLU function, etc. 4) The model sets a hidden layer, which contains n neurons; 5) The loss function can adopt the cross-entropy loss function (Cross-Entropy Loss); 6) The model can use the Adam optimization algorithm; 7) The number of model training times is set to n, and the number of training times is determined by the validation set monitoring method. The training is stopped when the validation set error no longer decreases to avoid overfitting or underfitting problems and improve the generalization ability of the model.

[0066] Step S14, calculating a first loss value between the output value and the abnormality recognition result, and adjusting the parameters in the initial model according to the first loss value through a back propagation algorithm to obtain an abnormality prediction model.

[0067] In this embodiment, the difference between the predicted output of the model and the actual abnormality identification result is first evaluated, and this difference is called the first loss value. Then, the weights and bias parameters of each layer in the neural network are adjusted according to this first loss value using the back propagation algorithm. The back propagation algorithm determines how to adjust these parameters to minimize the error by calculating the gradient of the loss function with respect to each parameter. This process is repeated until the prediction error of the model is reduced to an acceptable level or a preset number of training rounds is reached. In this way, the model can learn the ability to more accurately predict the activation of abnormal functions and the inactivation of abnormal functions, thereby obtaining a deviation prediction model.

[0068] Step S03: output corresponding control instructions according to the prediction results.

[0069] In this embodiment, the automatic parking function is turned on or off when the control instruction is executed. The operating state of the automatic parking function is determined by using the output of the abnormal prediction model, namely the probability of abnormal function activation and the probability of abnormal function inactivation. The prediction results of the model provide data support and decision-making basis for whether the state of the automatic parking function needs to be adjusted. Through intelligent prediction and control, the system can automatically adjust the automatic parking function according to real-time road conditions and vehicle status. When possible abnormal activation or inactivation is predicted, the system can control the opening or closing of the automatic parking function in advance to avoid potential safety risks caused by misoperation of the automatic parking function. At the same time, the system can adapt to different driving scenarios and road conditions to achieve personalization and optimization of the automatic parking function.

[0070] In some embodiments, as mentioned in step S03, the corresponding control instruction is output according to the prediction result, which may specifically be: When the probability of activation of the abnormal function exceeds a second threshold, outputting an automatic parking function shut-down instruction; When the probability that the function abnormality is not activated exceeds the third threshold, an automatic parking function activation instruction is output.

[0071] This embodiment provides a probability-based automated decision-making mechanism for intelligently controlling the automatic parking function. The abnormal activation probability and abnormal inactivation probability calculated by the abnormal prediction model are compared with the preset threshold to determine whether the state of the automatic parking function needs to be adjusted. When it is predicted that the automatic parking function has a high probability of abnormal activation, the system will output a shutdown command to prevent unnecessary activation. This may happen when the system predicts that an upcoming situation does not require the intervention of the automatic parking function, or to avoid driving discomfort caused by false activation.

[0072] On the contrary, when it is predicted that the automatic parking function has a high probability of abnormal inactivation, the system will output an activation command to ensure that the automatic parking function can work normally when needed, such as preventing the vehicle from rolling when parking on a slope. This probability-based control strategy improves the flexibility and adaptability of the automatic parking system, enabling it to respond more accurately to various driving scenarios, thereby improving driving safety and comfort. In this way, the automatic parking system can operate more intelligently, reduce the driver's operating burden, and avoid potential safety risks.

[0073] Based on the above embodiments, in some embodiments, the following steps may be performed to optimize the abnormality prediction model: Step S31, performing abnormality identification on the vehicle operation information to obtain real-time abnormality identification results.

[0074] In this embodiment, the system will perform abnormal identification on the acquired vehicle operation information in real time to obtain real-time abnormal identification results, which also include abnormal activation of functions and abnormal inactivation of functions to determine whether the state of the automatic parking function is abnormal. The purpose is to evaluate the accuracy of the abnormal prediction model based on the real-time abnormal identification results. The abnormal identification process is consistent with the abnormal identification process described in steps S21 to S23, and will not be repeated here.

[0075] Step S32, calculating a second loss value between the real-time anomaly recognition result and the prediction result, and optimizing the deviation prediction model according to the second loss value.

[0076] In this embodiment, the system compares the real-time anomaly recognition results with the prediction results of the model and calculates the difference between the two, that is, the second loss value. The second loss value reflects the deviation between the model prediction and the actual situation, and is an important indicator for evaluating the accuracy of the model prediction. By comparing the real-time anomaly recognition results with the prediction results of the model, the accuracy of the model prediction can be quantified. If the second loss value is large, it indicates that there is a large deviation in the model's prediction, and the model needs to be adjusted and optimized. This error analysis is an important part of the machine learning model training process and helps to improve the prediction performance of the model.

[0077] This embodiment allows the system to evaluate and monitor the performance of the anomaly prediction model. By analyzing the second loss value, the system can optimize the anomaly prediction model, thereby improving the accuracy of the model prediction.

[0078] Based on the above technical solution, the embodiment of the present invention obtains the first crystal oscillator frequency and environmental parameters of the clock source; then inputs the environmental parameters into a preset deviation prediction model to obtain a first deviation value output by the deviation prediction model; finally, the first crystal oscillator frequency is compensated and calibrated according to the first deviation value, and the time is tracked according to the compensation calibration result. The present invention significantly improves the accuracy and stability of the clock by real-time monitoring and compensating for the influence of environmental parameters on the crystal oscillator frequency. This dynamic compensation mechanism can adapt to environmental changes, reduce the influence of external factors on clock accuracy, and reduce the dependence on expensive hardware, thereby reducing the development and maintenance costs of the overall system. In addition, the present invention is implemented through software design, which is easy to integrate into the existing clock synchronization system, and the maintenance and update process is simple, without the need for complex hardware replacement or adjustment.

[0079] Please refer to Figure 3 , which is a flow chart of another method for controlling an automatic parking function provided by an embodiment of the present invention. In some embodiments, after executing step S03 to output a corresponding control instruction according to the prediction result, the following may also be executed: Figure 3 Steps shown: Step S41, when the control instruction is an automatic parking function activation instruction, it is determined according to the vehicle operation information whether the automatic parking function is in an activated state.

[0080] If not, execute step S42.

[0081] In this embodiment, upon receiving the automatic parking function activation instruction, the system first checks whether the automatic parking function is already activated according to the real-time operation information of the vehicle. If the automatic parking function is not activated (i.e., the system determines that the result is "no"), step S42 is executed, and the system actively activates the automatic parking function to ensure that the function can be activated when needed.

[0082] Furthermore, if the automatic parking function is already turned on (i.e., the system determines that the result is "yes"), no operation is required because the function is already in standby mode and can respond to changes in driving conditions at any time.

[0083] Step S42, turning on the automatic parking function.

[0084] Step S43, determining whether the automatic parking function is activated.

[0085] If yes, then go to step S44. If no, then go to step S45.

[0086] In this embodiment, after ensuring that the automatic parking function is in the standby state, the system checks whether the automatic parking function has been activated. If the automatic parking function has been activated, the system will execute step S44, waiting for the automatic parking function to exit after completing its task, and then the system will turn off the automatic parking function to avoid unnecessary energy consumption and potential misoperation. The automated shutdown process reduces the operating burden of the driver, allowing them to focus more on driving, thereby improving safety.

[0087] Step S44, after the automatic parking function is exited, turn off the automatic parking function.

[0088] Step S45, determining whether the time of receiving the automatic parking function closing instruction exceeds a fourth threshold.

[0089] If yes, execute step S46.

[0090] In this embodiment, when the automatic parking function is not activated, the system will evaluate whether the time from receiving the instruction to turn off the automatic parking function to the present has exceeded the preset fourth threshold. The fourth threshold is set to avoid the automatic parking function being in a passively turned-on state for a long time, resulting in unintended activation against the driver's intention.

[0091] If the time of receiving the automatic parking function shut-down instruction exceeds the fourth threshold, the system will execute step S46 to shut down the automatic parking function.

[0092] Furthermore, if the time of receiving the automatic parking function closing instruction does not exceed the fourth threshold, the system does not perform any operation and keeps the automatic parking function in the current state.

[0093] Step S46, turning off the automatic parking function.

[0094] Based on the above technical solution, this embodiment ensures that the automatic parking function can be activated quickly and accurately when needed when receiving the automatic parking function activation command. If the system finds that the automatic parking function has not been activated, it will perform the activation operation to ensure that the function is on standby at any time to respond to possible parking needs. This automated inspection and activation process not only improves the response speed, but also reduces the driver's operating burden, making the driving process smoother and safer. Through intelligent management, the accuracy and efficiency of the automatic parking function are improved, while the stability and safety of the system are also enhanced, ensuring that the best driving experience can be provided under various driving conditions.

[0095] Please refer to Figure 4 , which is a flow chart of another method for controlling an automatic parking function provided by an embodiment of the present invention. In some embodiments, after executing step S03 to output a corresponding control instruction according to the prediction result, the following may also be executed: Figure 4 Steps shown: Step S51, when the control instruction is an automatic parking function off instruction, it is determined according to the vehicle operation information whether the automatic parking function is in an off state.

[0096] If not, execute step S52.

[0097] In this embodiment, after receiving the instruction to turn off the automatic parking function, the system will immediately evaluate whether the automatic parking function is already in the off state according to the vehicle operation information. The purpose is to identify the current state of the automatic parking function, thereby avoiding unnecessary repeated operations. If the system confirms that the automatic parking function has not been turned off, step S52 is executed to turn off the automatic parking function.

[0098] This embodiment can ensure that the automatic parking function is turned off in time when it is not needed through clear shutdown instructions, avoiding unnecessary energy consumption and system load. The automated shutdown process improves the system's response speed to the shutdown instruction. When the automatic parking function is not needed, turning it off in time can prevent automatic parking at inappropriate times and improve driving safety.

[0099] Step S52, turning off the automatic parking function.

[0100] Step S53, determining whether the time of receiving the automatic parking function closing instruction exceeds a fifth threshold.

[0101] If yes, execute step S54.

[0102] In this embodiment, the system will evaluate whether the time from receiving the instruction to turn off the automatic parking function to the present has exceeded the preset fifth threshold. The setting of the fifth threshold helps to avoid the automatic parking function being in a passively turned off state for a long time, resulting in the situation that the automatic parking function is not activated unexpectedly against the driver's intention.

[0103] Step S54, turning on the automatic parking function.

[0104] Based on the above technical solution, this embodiment ensures that the automatic parking function can be turned off in time when it is not needed, thereby avoiding automatic parking at inappropriate times and improving driving safety and vehicle control accuracy.

[0105] Please refer to Figure 5 , is a schematic diagram of the structure of an automatic parking function control system provided by an embodiment of the present invention. Figure 5 As shown, the automatic parking function control system may include a sensor 1, a chassis controller 2, an actuator 3, a satellite positioning module 4, a network connection module 5, a vehicle controller 6 and a cloud computing platform 7, wherein: The sensor 1, the actuator 3, the satellite positioning module 4, the network connection module 5, the vehicle controller 6 and the cloud computing platform 7 are respectively connected to the chassis controller 2 in direct or indirect communication; The sensor 1, the network connection module 5, the satellite positioning module 4 and the vehicle controller 6 obtain the vehicle operation information; The cloud computing platform 7 trains the anomaly prediction model; Actuator 3 executes the control instruction; The chassis controller 2 includes a processor and a memory connected to the processor for communication. The memory is used to store at least one instruction. When the instruction is loaded and executed by the processor, the method for controlling the automatic parking function as described above is implemented. The execution method and beneficial effects are similar and will not be repeated here.

[0106] In some embodiments, sensor 1 is responsible for collecting vehicle operation information, which may specifically include a wheel speed sensor, a master cylinder pressure sensor, an acceleration sensor, a steering wheel angle sensor, and the like.

[0107] The chassis controller 2 is responsible for collecting sensor information, controlling actuator actions, transmitting sensor information and actuator action status to the networking module 5, and receiving and executing control instructions of the networking module 5. The chassis controller 2 retains the integrated automatic parking function control program to ensure the function response speed while avoiding functional failure caused by failure of the networking module 5. Actions that do not require high response time are directly controlled by the control instructions of the networking module 5. The automatic parking control software in the chassis controller 2 can be quickly updated and optimized according to the control instructions of the networking module 5.

[0108] The actuator 3 is responsible for the braking force control and the parking force control, and the specific hardware structure is not limited.

[0109] The satellite positioning module 4 is responsible for providing the real-time position information of the vehicle.

[0110] The network connection module 5 is responsible for collecting chassis controller information, vehicle controller information, and satellite positioning information and sending them to the cloud computing platform 7 , while collecting control information from the cloud computing platform 7 and sending it to the chassis controller 2 .

[0111] The vehicle controller 6 is responsible for collecting vehicle information related to the automatic parking function and providing it to the network connection module 5, such as vehicle torque, driver status, voice requirements, etc.

[0112] The cloud computing platform 7 uses the collected data to train the automatic parking function control program and sends control instructions to the network module 5. The cloud computing platform 7 identifies the road condition information and unexpected function recognition from the information collected by the network module 5, and continuously optimizes the automatic parking function control program through intelligent software control logic to realize the intelligence of the automatic parking function. The control instructions sent by the cloud computing platform 7 include two parts: actuator action control instructions and basic program optimization instructions. The actuator action control instructions are used to directly control the action of the actuator 3, and the basic program optimization instructions are used to quickly update the automatic parking basic control program.

[0113] This embodiment provides an electronic device, including a processor and a memory, the memory is used to store at least one instruction, and when the instruction is loaded and executed by the processor, the method for controlling the automatic parking function described above is implemented. The execution method and beneficial effects are similar and will not be repeated here.

[0114] An embodiment of the present invention provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the method for controlling the automatic parking function described above is implemented. The execution method and beneficial effects thereof are similar and will not be described in detail here.

[0115] It should be noted that although the above describes the various steps in a specific order, it does not mean that the various steps must be executed in the above specific order. In fact, some of these steps can be executed concurrently or even in a different order as long as the required functions can be achieved.

[0116] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for controlling an automatic parking function, characterized in that: include: Obtain vehicle operation information; The vehicle operation information includes road condition information and vehicle control information; Inputting the vehicle operation information into a preset abnormality prediction model to obtain a prediction result output by the abnormality prediction model; the prediction result includes a probability of function abnormality activation and a probability of function abnormality non-activation; A corresponding control instruction is output according to the prediction result; when the control instruction is executed, the automatic parking function is controlled to be turned on or off.

2. The method according to claim 1, characterized in that The process of establishing the abnormal prediction model includes: Acquire historical vehicle operation information, and perform abnormality identification on the historical vehicle operation information to obtain abnormality identification results; the abnormality identification results include abnormal function activation and abnormal function inactivation; Performing data cleaning and formatting on the historical vehicle operation information, and constructing a training data set based on the processed historical vehicle operation information; Establishing an initial model, and inputting the training data set into the initial model for forward propagation processing to obtain an output value; A first loss value between the output value and the abnormality recognition result is calculated, and the parameters in the initial model are adjusted according to the first loss value through a back propagation algorithm to obtain the abnormality prediction model.

3. The method according to claim 2, characterized in that The performing abnormality identification on the historical vehicle operation information to obtain an abnormality identification result includes: determining status information of an automatic parking function according to the historical vehicle operation information; When the automatic parking function is in an activated state and satisfies a first preset condition, determining that the abnormality identification result is abnormal activation of the function; The first preset condition includes: receiving an input automatic parking function shut-off signal; and / or the duration of the automatic parking function being in an activated state is less than a first threshold; and / or the automatic parking function is repeatedly activated within a set duration; When the automatic parking function is in an inactivated state and meets a second preset condition, determining that the abnormality identification result is that the function is abnormal and inactivated; The second preset condition includes: the automatic parking function directly jumps to the activation state after receiving an input automatic parking function activation signal; and / or, the automatic parking function activation signal is received after the vehicle is powered on and before starting.

4. The method according to claim 3, characterized in that The method further comprises: Performing abnormality identification on the vehicle operation information to obtain real-time abnormality identification results; A second loss value between the real-time anomaly identification result and the prediction result is calculated, and the deviation prediction model is optimized according to the second loss value.

5. The method according to claim 1, characterized in that Outputting a corresponding control instruction according to the prediction result includes: When the probability of activation of the abnormal function exceeds a second threshold, outputting an automatic parking function shut-down instruction; When the function abnormality inactivation probability exceeds a third threshold, an automatic parking function activation instruction is output.

6. The method according to claim 5, characterized in that After outputting the corresponding control instruction according to the prediction result, the method further includes: When the control instruction is the automatic parking function activation instruction, judging whether the automatic parking function is in the activation state according to the vehicle operation information; If not, turn on the automatic parking function; determining whether the automatic parking function is activated; If yes, then after the automatic parking function exits, turn off the automatic parking function; If not, determining whether the time of receiving the automatic parking function closing instruction exceeds a fourth threshold; If exceeded, the automatic parking function will be turned off.

7. The method according to claim 5, characterized in that After outputting the corresponding control instruction according to the prediction result, the method further includes: When the control instruction is the automatic parking function off instruction, judging whether the automatic parking function is in an off state according to the vehicle operation information; If not, turn off the automatic parking function; determining whether the time of receiving the automatic parking function shut-down instruction exceeds a fifth threshold; If it exceeds, the automatic parking function will be turned on.

8. A system for controlling an automatic parking function, characterized in that: It includes sensors, chassis controllers, actuators, satellite positioning modules, networking modules, vehicle controllers and cloud computing platforms, including: The sensor, the actuator, the satellite positioning module, the network connection module, the vehicle controller and the cloud computing platform are respectively connected to the chassis controller directly or indirectly for communication; The sensor, the network connection module, the satellite positioning module and the vehicle controller obtain vehicle operation information; The cloud computing platform trains an anomaly prediction model; The actuator executes the control instruction; The chassis controller includes a processor and a memory communicatively connected to the processor, wherein the memory is used to store at least one instruction, and when the instruction is loaded and executed by the processor, the method for controlling the automatic parking function as described in any one of claims 1-7 is implemented.

9. An electronic device, characterized in that: include: A processor and a memory, wherein the memory is used to store at least one instruction, and when the instruction is loaded and executed by the processor, the method for controlling the automatic parking function as described in any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for controlling an automatic parking function as claimed in any one of claims 1 to 7 is implemented.

Images