A method for estimating delay of a drive-by-wire system for an autonomous driving vehicle
The delay estimation of the throttle-braking system and steering system of an autonomous vehicle is solved by the function determined by the minimum mean square error method, which solves the problem of insufficient delay estimation and improves system performance and reliability.
Patent Information
- Application Number
- CN202210296000.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-03-24
- Publication Date
- 2024-11-15
- Estimated Expiration
- 2042-03-24
AI Technical Summary
The self-driving vehicle line control execution system has insufficient in estimating the delay of the throttle-braking system and steering system, resulting in problems such as slow start, slow acceleration, slow braking and insufficient steering.
The minimum mean square error method is used to determine the normalized step size calculation function, error calculation function and weight iterative function. Through these functions, delay estimation of the dispatched data and feedback data of the actuator can be used to process the throttle-braking system and steering at the same time. The system's delay information and save the delay data under different working conditions.
It effectively solves the delay estimation problem of the line-controlled execution system, reduces the difficulty of delay estimation, and increases the storage and analysis functions of delay data, improving the performance and reliability of the system.
Smart Images

Figure CN114684195B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of data processing technology, and in particular to a delay estimation method for a control-by-wire execution system of an autonomous driving vehicle. Background Art
[0002] The control-by-wire execution system of an autonomous driving vehicle is generally divided into a throttle-brake system actuator and a steering system actuator. For the throttle-brake system actuator of a fuel vehicle or electric vehicle, the output of the engine or drive motor has a pure delay effect until the vehicle accelerates; the process of the brake system building up pressure in the pressure master cylinder also has a certain pure delay effect. If the control-by-wire execution system has no estimation capability for the delay of the throttle-brake system actuator, problems such as slow starting, slow acceleration, and slow braking will occur. For the steering system actuator, there will also be a certain pure delay effect due to the transmission of the steering force mechanical structure. If the control-by-wire execution system has no estimation capability for the delay of the steering system actuator, problems such as understeering will occur when the vehicle turns a right angle. Summary of the invention
[0003] The purpose of the present invention is to address the defects of the prior art and provide a method for estimating the delay of the control-by-wire execution system of an autonomous driving vehicle, an electronic device, and a computer-readable storage medium. After setting the delay estimation model function, the corresponding normalized step calculation function, error calculation function, and weight iteration function are determined based on the minimum mean square error (MMSE) method, and the delay of the actuator's sent data and feedback data is estimated based on the above functions. Through the present invention, the same set of functions can be used to estimate the delay of the throttle-brake system actuator and the steering system actuator, and the delay information of different actuators under different working conditions can also be saved; in this way, not only can the delay estimation problem of the control-by-wire execution system be solved, but also the difficulty of delay estimation can be reduced.
[0004] To achieve the above-mentioned purpose, a first aspect of an embodiment of the present invention provides a method for estimating delay of a drive-by-wire execution system of an autonomous driving vehicle, the method comprising:
[0005] Acquire first identification data;
[0006] For the actuator corresponding to the first identification data, the real-time working condition information, the actuator sending data and the actuator feedback data at the current time t are collected to generate the corresponding first working condition data and the first sending data s t and the first feedback data f t ;
[0007] The first vector S is formed by the latest specified number L of first sent data. t ; The first vector S t for (s t ,st-1 …s t-(i-1) …s t-(L-1) ),1≤i≤L;
[0008] The first vector S t Substitute the predetermined step length calculation function to calculate the corresponding first step length data sp t ;
[0009] The first vector S t and the first feedback data f t Substitute the predetermined error calculation function to calculate the corresponding first error data err t ;
[0010] The first vector S t , the first step length data sp t and the first error data err t Substitute the predetermined weight iteration function to calculate the latest first delay weight vector B t+1 ; The first delay weight vector B t+1 for (b t+1(1) ,b t+1(2) …b t+1(i) …b t+1(L) ), b t+1(i) is the first delay weight vector B t+1 The delay weight vector coefficient of ;
[0011] From the first delay weight vector B t+1 (b t+1(2) …b t+1(i) …b t+1(L) ), select the maximum value as the maximum delay weight coefficient b at the current time t t(max) ;
[0012] According to the maximum delay weight coefficient b t(max) and the preset acquisition frequency T s Calculate the first delay data Td at the current time t t , Td t =T s *b t(max) ;
[0013] The first identification data, the first working condition data, and the first sent data s t , the first feedback data f t and the first delay data Td t The delay estimation data set R at the current time t t and save it.
[0014] Preferably, the wire control execution system includes a throttle-brake system actuator and a steering system actuator; correspondingly, the first identification data includes a throttle-brake system actuator identification and a steering system actuator identification;
[0015] When the first identification data is an accelerator-brake system actuator identification, the first operating condition data includes a starting condition, a parking condition, a sudden acceleration condition, a sudden braking condition, a slow acceleration condition and a slow braking condition, the actuator-issued data is a longitudinal control-issued acceleration data, and the actuator feedback data is a vehicle feedback equivalent acceleration data;
[0016] When the first identification data is the steering system actuator identification, the first operating condition data includes a small turning condition to the left, a small turning condition to the right, a large turning condition to the left, a large turning condition to the right, a right-angle turn condition to the left, and a right-angle turn condition to the right. The data sent down by the actuator is the sent steering angle percentage data of the lateral control, and the actuator feedback data is the vehicle feedback equivalent steering angle percentage data.
[0017] Preferably, before acquiring the first identification data, determining the error calculation function, the weight iteration function and the step length calculation function specifically includes:
[0018] Step 31, determine the delay estimation model function as:
[0019]
[0020] in, is the predicted feedback data at time n, x n-(i-1) is the data sent by the L actuators most recently collected up to time n, X n For x n-(i-1) The sent data vector is composed of For X n The transposed vector, W n is the delay weight vector at time n, the delay weight vector W n is (w n(1) ,w n(2) …w n(i) …w n(L) ), w n(i) is the delay weight vector W n The delay weight vector coefficient of ;
[0021] Step 32, determining the error calculation function at time n according to the delay estimation model function:
[0022]
[0023] Among them, e n is the error data at time n, yn is the actuator feedback data collected at time n;
[0024] Step 33, based on the minimum mean square error MMSE criterion, the minimization objective function at time n is determined according to the error calculation function:
[0025]
[0026] Among them, J n (W) is the minimization objective function with the delay weight vector W as the independent variable;
[0027] Step 34, determining the weight iteration function for weight vector iteration at time n according to the minimization objective function:
[0028]
[0029] Among them, μ n is the step length data at time n;
[0030] Step 35, setting the step length in the weight iteration function to the normalized step length, and determining the step length calculation function to be:
[0031] Among them, α and β are preset constants, 0<α<2, 0≤β, x n Send data to the actuator collected at time n.
[0032] Preferably, the first vector S t Substitute the predetermined step length calculation function to calculate the corresponding first step length data sp t , specifically including:
[0033] Assume preset constants α=1, β=0.00001;
[0034] For the first vector S t Transpose to get the corresponding first transposed vector
[0035] The preset constants α and β, the first transposed vector The first vector S t The first sent data s t , substitute into the step length calculation function Get the corresponding first step length data sp n for:
[0036]
[0037] Preferably, the first vector S t and the first feedback data ft Substitute the predetermined error calculation function to calculate the corresponding first error data err t , specifically including:
[0038] Get the delay weight vector B obtained at the previous time n-1 t ;
[0039] For the delay weight vector B t Transpose to get the corresponding delay weight transpose vector
[0040] Transpose the delay weight vector The first vector S t and the first feedback data f t Substitute into the error calculation function Get the corresponding first error data err t for:
[0041]
[0042] Preferably, the first vector S t , the first step length data sp t and the first error data err t Substitute the predetermined weight iteration function to calculate the latest first delay weight vector B t+1 , specifically including:
[0043] Get the delay weight vector B obtained at the previous time n-1 t ;
[0044] And the delay weight vector B t , the first vector S t , the first step length data sp t and the first error data err t Substitute into the weight iteration function W n+1 =W n +μ n e n X n Get the corresponding first delay weight vector B t+1 For: B t+1 =B t +sp t err t S t .
[0045] Preferably, the method further comprises:
[0046] A plurality of delay estimation data sets R stored according to the executor identification data pairst Perform a query and draw the delay curve of the corresponding actuator under different working conditions based on the query results.
[0047] Preferably, the method further comprises:
[0048] A plurality of delay estimation data groups R stored according to actuator identification data and operating condition data t Perform a query and draw the delay curve of the corresponding actuator under the specified working conditions based on the query results.
[0049] A second aspect of an embodiment of the present invention provides an electronic device, including: a memory, a processor, and a transceiver;
[0050] The processor is used to couple with the memory, read and execute instructions in the memory, so as to implement the method steps described in the first aspect above;
[0051] The transceiver is coupled to the processor, and the processor controls the transceiver to send and receive messages.
[0052] A third aspect of an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer instructions. When the computer instructions are executed by a computer, the computer executes the instructions of the method described in the first aspect above.
[0053] The embodiment of the present invention provides a method for estimating the delay of the control-by-wire execution system of an autonomous driving vehicle, an electronic device, and a computer-readable storage medium. After setting the delay estimation model function, the corresponding normalized step calculation function, error calculation function, and weight iteration function are determined based on the minimum mean square error method, and the delay of the actuator's sent data and feedback data is estimated based on the above functions. Through the present invention, the same set of functions can be used to estimate the delay of the throttle-brake system actuator and the steering system actuator, and the delay information of different actuators under different working conditions can be saved; it not only solves the delay estimation problem of the control-by-wire execution system and reduces the difficulty of delay estimation, but also increases the storage and analysis functions of delay data. BRIEF DESCRIPTION OF THE DRAWINGS
[0054] Figure 1 A schematic diagram of a delay estimation method for a control-by-wire execution system of an autonomous driving vehicle provided in Embodiment 1 of the present invention;
[0055] Figure 2 A schematic diagram of the structure of an electronic device provided in Embodiment 2 of the present invention. DETAILED DESCRIPTION
[0056] In order to make the purpose, technical solution and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the 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.
[0057] Embodiment 1 of the present invention provides a method for estimating delay of a by-wire control execution system of an autonomous driving vehicle; the by-wire control execution system of an autonomous driving vehicle to which the method is applicable includes a throttle-brake system actuator and a steering system actuator.
[0058] The wire control execution system in the method determines a set of functions (error calculation function, weight iteration function and step length calculation function) for delay estimation based on the delay estimation model function before performing delay estimation. The process of determining the error calculation function, weight iteration function and step length calculation function in advance includes the following steps:
[0059] Step A1, determining the delay estimation model function as:
[0060]
[0061] in, is the predicted feedback data at time n, x n-(i-1) is the data sent by the L actuators most recently collected up to time n, X n For x n-(i-1) The sent data vector is composed of For X n The transposed vector, W n is the delay weight vector at time n, the delay weight vector W n is (w n(1) ,w n(2) …w n(i) …w n(L) ), w n(i) is the delay weight vector W n The delay weight vector coefficient of ;
[0062] Here, the delay estimation model function is actually constructed based on the delay characteristics of signal transmission; if the observation window length is T max , the signal acquisition frequency is T s , then L = T max / T s ;
[0063] Step A2, determining the error calculation function at time n according to the delay estimation model function:
[0064]
[0065] Among them, e n is the error data at time n, y n is the actuator feedback data collected at time n;
[0066] Here, the error is the difference between the predicted value and the actual observed value;
[0067] Step A3, based on the minimum mean square error criterion, the minimization objective function at time n is determined according to the error calculation function:
[0068]
[0069] Among them, J n (W) is the minimization objective function with the delay weight vector W as the independent variable;
[0070] Step A4, determining the weight iteration function for weight vector iteration at time n according to the minimization objective function:
[0071]
[0072] Among them, μ n is the step length data at time n;
[0073] Here, the delay weight vector W1 is an initial setting vector, and each time n can be recursively obtained based on the initial setting vector. At each time n, the delay weight vector of the next time n+1 can be estimated based on the minimum mean square error criterion;
[0074] Step A5, set the step length in the weight iteration function to the normalized step length, and determine the step length calculation function as:
[0075] Among them, α and β are preset constants, 0<α<2, 0≤β, x n Send data to the actuator collected at time n.
[0076] Figure 1 FIG. 1 is a schematic diagram of a delay estimation method for a wire control execution system of an autonomous driving vehicle provided in Embodiment 1 of the present invention. After determining the above set of functions, the wire control execution system can be as follows: Figure 1 The same set of functions is used to estimate the delay of the system including the throttle-brake system actuator and the steering system actuator; the method mainly includes the following steps:
[0077] Step 1, obtaining first identification data;
[0078] The first identification data includes an accelerator-brake system actuator identification and a steering system actuator identification.
[0079] Here, when the first identification data is an identification of an accelerator-brake system actuator, the wire control actuation system performs delay estimation on the accelerator-brake system actuator, and when the first identification data is an identification of a steering system actuator, delay estimation is performed on the steering system actuator.
[0080] Step 2: For the actuator corresponding to the first identification data, the real-time working condition information, the actuator sent data and the actuator feedback data at the current time t are collected to generate the corresponding first working condition data and the first sent data s t and the first feedback data f t ;
[0081] Among them, when the first identification data is the throttle-brake system actuator identification, the first operating condition data includes starting condition, parking condition, sudden acceleration condition, sudden braking condition, slow acceleration condition and slow braking condition, the data sent by the actuator is the acceleration data sent for longitudinal control, and the actuator feedback data is the vehicle feedback equivalent acceleration data; when the first identification data is the steering system actuator identification, the first operating condition data includes a small turning condition to the left, a small turning condition to the right, a large turning condition to the left, a large turning condition to the right, a right-angle turn to the left and a right-angle turn to the right, the data sent by the actuator is the steering angle percentage data sent for lateral control, and the actuator feedback data is the vehicle feedback equivalent steering angle percentage data.
[0082] Here, when the first identification data is the throttle-brake system actuator identification, the delay estimation of the throttle-brake system actuator is performed. At this time, the first working condition data obtained is information related to the working conditions of the throttle-brake system actuator: the starting working condition indicates that the vehicle is just starting, the acceleration is positive and increases from small to large, and the vehicle speed increases from 0; the parking working condition indicates that the vehicle is in a parking state, and the acceleration and speed are close to 0; the rapid acceleration working condition indicates that the acceleration of the vehicle is positive and greater than a rapid acceleration threshold, which is 2m / s by default. 2 ; The emergency braking condition indicates that the vehicle's acceleration is negative and less than a sudden deceleration threshold, which defaults to -2m / s 2 ; The slow acceleration condition indicates that the vehicle's acceleration is positive and remains within a small acceleration range, which is less than or equal to 1m / s by default. 2 ; Slow braking condition means that the vehicle's acceleration is negative and remains within a specified acceleration range, which defaults to greater than or equal to -1m / s 2In addition, the actuator data collected at this time is the acceleration data of the longitudinal control, that is, the acceleration information corresponding to the throttle-brake control command issued by the throttle-brake system actuator; the actuator feedback data collected is the vehicle feedback equivalent acceleration data, and the collection method of the equivalent acceleration data is: obtaining the real-time displacement information of the vehicle from the vehicle positioning system, obtaining the vehicle speed information before and after the displacement from the vehicle chassis module, and calculating the corresponding equivalent acceleration information based on the time, vehicle speed and displacement information before and after the displacement.
[0083] When the first identification data is the steering system actuator identification, the delay estimation of the steering system actuator is performed. At this time, the first operating condition data obtained is information related to each operating condition of the steering system actuator: the left / right small steering condition indicates that the vehicle makes a small turn to the left / right, and the turning angle is less than a specified threshold for a small turn angle; the left / right large steering condition indicates that the vehicle makes a large turn to the left / right, and the turning angle is greater than a specified threshold for a large turn angle, and the specified threshold for the large turn angle is greater than the specified threshold for the small turn angle; the left / right right-angle turn condition indicates that the vehicle makes a right-angle turn to the left / right, and the turning angle is close to or equal to 90°. In addition, the actuator-issued data collected at this time is the angle percentage data for the lateral control, that is, the angle percentage information corresponding to the steering wheel control command issued by the steering system actuator. The value of the angle percentage data sent down is between [-1,1]; the actuator feedback data collected is the vehicle feedback equivalent angle percentage data, and the equivalent angle percentage data is collected as follows: the vehicle positioning system detects the vehicle's real-time yaw rate and speed to calculate the corresponding road curvature, and after obtaining the road curvature from the vehicle positioning system, the corresponding front wheel turning angle is calculated according to the front wheel turning angle = atan (road curvature * vehicle wheelbase), and then The equivalent rotation angle percentage data is obtained, and the value of the equivalent rotation angle percentage data is also between [-1,1].
[0084] Step 3: The first vector S is formed by combining the most recent specified number L of first sent data t ;
[0085] Among them, the first vector S t for (s t ,s t-1 …s t-(i-1) …s t-(L-1) ),1≤i≤L.
[0086] Step 4: transform the first vector S t Substitute the predetermined step length calculation function to calculate the corresponding first step length data sp t ;
[0087] Specifically, it includes: step 41, setting preset constants α=1, β=0.00001;
[0088] Here, α can also be set to other values between 0 and 2, and β can also be set to other values greater than or equal to zero, but in order to prevent the step size data from being too large, β is often set to a decimal greater than zero;
[0089] Step 42, for the first vector S t Transpose to get the corresponding first transposed vector
[0090] Step 43, set the preset constants α and β, the first transposed vector The first vector S t The first data sent in t , substitute into the step length calculation function Get the corresponding first step length data sp n for:
[0091]
[0092] Here, the first transposed vector As The first vector S t The first data sent in t As x n , and then convert the preset constants α and β into constants, we can get the first step length data sp n .
[0093] Step 5: transform the first vector S t and the first feedback data f t Substitute the predetermined error calculation function to calculate the corresponding first error data err t ;
[0094] Specifically, step 51 is to obtain the delay weight vector B obtained at the previous time n-1. t ;
[0095] Here, it can be seen from the previous text that when the initial delay weight vector B1 is known, the delay weight vector B at any time can be deduced step by step;
[0096] Step 52: Delay weight vector B t Transpose to get the corresponding delay weight transpose vector
[0097] Step 53, transpose the delay weight vector The first vector S t and the first feedback data f t Substitute into the error calculation function Get the corresponding first error data err t for:
[0098] Here, the first feedback data f t As y n , transpose the delay weight vector As The first vector S t As X n , we can get the first error data err t .
[0099] Step 6: Transform the first vector S t , the first step is long data sp t and the first error data err t Substitute the predetermined weight iteration function to calculate the latest first delay weight vector B t+1 ;
[0100] Among them, the first delay weight vector B t+1 for (b t+1(1) ,b t+1(2) …b t+1(i) …b t+1(L) ), b t+1(i) is the first delay weight vector B t+1 The delay weight vector coefficient of ;
[0101] Specifically, step 61, obtaining the delay weight vector B obtained at the previous time n-1 t ;
[0102] Step 62: Delay weight vector B t , the first vector S t , the first step is long data sp t and the first error data err t Substitute into the weight iteration function W n+1 =W n +μ n e n X n Get the corresponding first delay weight vector B t+1 For: B t+1 =B t +sp t err t S t .
[0103] Here, the delay weight vector B t As W n , the first vector S t As Xn , the first step length data sp t As μ n , the first error data err t As e n , the first delay weight vector B can be calculated t+1 .
[0104] Step 7: From the first delay weight vector B t+1 (b t+1(2) …b t+1(i) …b t+1(L) ), select the maximum value as the maximum delay weight coefficient b at the current time t t(max) .
[0105] Step 8: According to the maximum delay weight coefficient b t(max) and the preset acquisition frequency T s Calculate the first delay data Td at the current time t t , Td t =T s *b t(max) .
[0106] It should be noted that, in the first embodiment of the present invention, a delay alarm threshold and an alarm quantity threshold may be preset, and the first delay data Td is obtained by the current step in the online linear control system. t The first delay data Td is processed based on the delay alarm threshold t Make a judgment; if the first delay data Td t If the preset delay alarm threshold is exceeded, the first identification data, the first working condition data, and the first sent data s at the current time t are used. t , the first feedback data f t and the first delay data Td t The corresponding first alarm data group is formed and saved with the delay alarm threshold; the number of first alarm data groups generated within the most recent specified time period is counted to generate the corresponding first alarm quantity, and when the first alarm quantity exceeds the alarm quantity threshold, the corresponding delay too long alarm processing is performed. There are many ways to implement the delay too long alarm processing, which can be set by the specific project implementation, and will not be further described here.
[0107] Step 9, the first identification data, the first working condition data, the first sent data s t , the first feedback data f t and the first delay data Td t The delay estimation data set R at the current time t t and save it.
[0108] Here, after completing the delay estimation of the accelerator-brake system actuator or the steering system actuator, the first embodiment of the present invention will also include the first identification data, the first working condition data, and the first sent data s t , the first feedback data f t and the first delay data Td t The delay estimation data set R t Stored for subsequent query, call and analysis.
[0109] In summary, if the linear control system continuously executes the above steps 1-9 for the throttle-brake system actuator and the steering system actuator from time t=0, the real-time tracking of the delay characteristics of the two actuators can be achieved. During real-time tracking, because the collected feedback data is actually the delayed feedback information of the data sent at a certain time before, there will be no first feedback data f in the initial period starting from t=0. t and the first delay data Td t , until the first feedback data f t After it appears, the corresponding delay estimation information, that is, the first delay data Td, can be continuously obtained. t In addition, the linear control system can also establish a sliding window monitoring mechanism for the actuator delay characteristics of the throttle-brake system actuator and the steering system actuator through the above steps 1-9, that is, in the working process of the two actuators, except for the above initial period of time without delayed feedback, a period of time is randomly selected as a sliding window, and the actuator delay characteristics in the current sliding window are estimated based on the sent data and feedback data collected within the sliding window time period and an initialized delay weight vector.
[0110] In addition, based on the stored multiple delay estimation data sets R t Embodiment 1 of the present invention can also provide a variety of query mechanisms for the linear control system, such as query by actuator identification data, query by time, query by actuator identification data and operating condition data, and query by a combination of actuator identification, time and operating condition data.
[0111] Furthermore, the first embodiment of the present invention can also provide a characteristic curve drawing processing method for a linear control system based on the above query means.
[0112] One of the processing methods for drawing characteristic curves by querying is: multiple delay estimation data sets R stored according to the actuator identification data t The query is performed, and the delay curve of the corresponding actuator under different working conditions is drawn according to the query result, specifically: receiving the actuator identification keyword; and according to the actuator identification keyword, the multiple delay estimation data groups R stored tPolling is performed and one or more delay estimation data sets R that match the first identification data with the executor identification key are matched. t The extracted data is recorded as the first data set; and the data set R is estimated according to each delay t The first operating condition data is divided into one or more second data group sets, so that the delay estimation data group R in each second data group set t The first working condition data of the delay estimation data group R in each second data group set are kept consistent; and the delay estimation data group R is calculated in chronological order. t Sort and extract the sorted delay estimation data set R t The first data sent t , the first feedback data f t and the first delay data Td t Draw the corresponding time curve of sent data, time curve of feedback data and time curve of delay change.
[0113] In this way, the delay characteristics of the accelerator-brake system actuator and the steering system actuator under various working conditions can be intuitively detected through the above-mentioned characteristic curve drawing processing method.
[0114] Another processing method for drawing characteristic curves by querying is: according to the actuator identification data and the working condition data, multiple delay estimation data groups R are saved. t The query is performed, and the delay curve of the corresponding actuator under the specified working condition is drawn according to the query result, specifically: receiving the actuator identification keyword and the working condition keyword; and according to the actuator identification keyword and the working condition keyword, the multiple delay estimation data groups R stored are t Polling is performed, and one or more delay estimation data sets R whose first identification data matches the actuator identification keyword and whose first operating condition data matches the operating condition keyword are received. t Extracted and recorded as a third data set; and in the third data set, the delay estimation data set R is sorted in chronological order. t Sort and extract the sorted delay estimation data set R t The first data sent t , the first feedback data f t and the first delay data Td t Draw the corresponding time curve of sent data, time curve of feedback data and time curve of delay change.
[0115] In this way, the delay characteristics of the accelerator-brake system actuator and the steering system actuator under the specified working condition can be intuitively detected through the above characteristic curve drawing processing method.
[0116] In addition, in the drawing process of the above two characteristic curves, if the number of the first alarm data groups mentioned above is not 0, the first embodiment of the present invention can also mark the time points corresponding to each first alarm data group as abnormal points on the drawn downlink data time curve, feedback data time curve and delay change time curve.
[0117] Figure 2 This is a schematic diagram of the structure of an electronic device provided in the second embodiment of the present invention. The electronic device may be the aforementioned terminal device or server, or may be a terminal device or server connected to the aforementioned terminal device or server to implement the method of the embodiment of the present invention. Figure 2 As shown, the electronic device may include: a processor 301 (such as a CPU), a memory 302, and a transceiver 303; the transceiver 303 is coupled to the processor 301, and the processor 301 controls the transceiver 303. Various instructions may be stored in the memory 302 to complete various processing functions and implement the processing steps described in the aforementioned method embodiment. Preferably, the electronic device involved in the embodiment of the present invention also includes: a power supply 304, a system bus 305 and a communication port 306. The system bus 305 is used to realize the communication connection between components. The above-mentioned communication port 306 is used for connecting and communicating between the electronic device and other peripherals.
[0118] exist Figure 2 The system bus 305 mentioned in the figure can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, etc. The system bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus. The communication interface is used to realize the communication between the database access device and other devices (such as clients, read-write libraries, and read-only libraries). The memory may include random access memory (RAM), and may also include non-volatile memory (Non-Volatile Memory), such as at least one disk storage.
[0119] The above-mentioned processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), a graphics processing unit (GPU), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.
[0120] It should be noted that an embodiment of the present invention further provides a computer-readable storage medium, in which instructions are stored. When the computer-readable storage medium is run on a computer, the computer executes the method and processing process provided in the above embodiments.
[0121] An embodiment of the present invention further provides a chip for executing instructions, wherein the chip is used to execute the processing steps described in the aforementioned method embodiment.
[0122] The embodiment of the present invention provides a method for estimating the delay of a control-by-wire execution system of an autonomous driving vehicle, an electronic device, and a computer-readable storage medium. After setting the delay estimation model function, the corresponding normalized step calculation function, error calculation function, and weight iteration function are determined based on the minimum mean square error method, and the delay of the actuator's sent data and feedback data is estimated based on the above functions. Through the present invention, the same set of functions can be used to estimate the delay of the throttle-brake system actuator and the steering system actuator, and the delay information of different actuators under different working conditions can be saved, and the delay curves of different actuators under different working conditions can be observed, and abnormal delays of different actuators under different working conditions can be processed; it not only solves the delay estimation problem of the control-by-wire execution system and reduces the difficulty of delay estimation, but also increases the storage and analysis functions of delay data.
[0123] The professionals should further realize that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in the above description according to the function. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.
[0124] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.
[0125] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of 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 estimating delay of a drive-by-wire system of an autonomous driving vehicle, characterized in that: The method comprises: Acquire first identification data; For the actuator corresponding to the first identification data, the real-time working condition information, the actuator sending data and the actuator feedback data at the current time t are collected to generate the corresponding first working condition data and the first sending data s t and the first feedback data f t ; The first vector S is formed by the latest specified number L of first sent data. t ; The first vector S t for (s t ,s t-1 …s t-(i-1) …s t-(L-1) ),1≤i≤L; The first vector S t Substitute the predetermined step length calculation function to calculate the corresponding first step length data sp t ; The first vector S t and the first feedback data f t Substitute the predetermined error calculation function to calculate the corresponding first error data err t ; The first vector S t , the first step length data sp t and the first error data err t Substitute the predetermined weight iteration function to calculate the latest first delay weight vector B t+1 ; The first delay weight vector B t+1 for (b t+1(1) ,b t+1(2) …b t+1(i) …b t+1(L) ), b t+1(i) is the first delay weight vector B t+1 The delay weight vector coefficient of ; From the first delay weight vector B t+1 (b t+1(2) …b t+1(i) …b t+1(L) ), select the maximum value as the maximum delay weight coefficient b at the current time t t(max) ; According to the maximum delay weight coefficient b t(max) and the preset acquisition frequency T s Calculate the first delay data Td at the current time t t , Td t =T s *b t(max) ; The first identification data, the first working condition data, and the first sent data s t , the first feedback data f t and the first delay data Td t The delay estimation data set R at the current time t t and save it.
2. The delay estimation method of the automatic driving vehicle drive-by-wire system according to claim 1, characterized in that: The wire control execution system includes a throttle-brake system actuator and a steering system actuator; correspondingly, the first identification data includes a throttle-brake system actuator identification and a steering system actuator identification; When the first identification data is an accelerator-brake system actuator identification, the first operating condition data includes a starting condition, a parking condition, a sudden acceleration condition, a sudden braking condition, a slow acceleration condition and a slow braking condition, the actuator-issued data is a longitudinal control-issued acceleration data, and the actuator feedback data is a vehicle feedback equivalent acceleration data; When the first identification data is the steering system actuator identification, the first operating condition data includes a small turning condition to the left, a small turning condition to the right, a large turning condition to the left, a large turning condition to the right, a right-angle turn condition to the left, and a right-angle turn condition to the right. The data sent down by the actuator is the sent steering angle percentage data of the lateral control, and the actuator feedback data is the vehicle feedback equivalent steering angle percentage data.
3. The delay estimation method of the automatic driving vehicle drive-by-wire system according to claim 1, characterized in that: Before obtaining the first identification data, determining the error calculation function, the weight iteration function and the step length calculation function specifically includes: Step 31, determine the delay estimation model function as: in, is the predicted feedback data at time n, x n-(i-1) is the data sent by the L actuators most recently collected up to time n, X n For x n-(i-1) The sent data vector is composed of For X n The transposed vector, W n is the delay weight vector at time n, the delay weight vector W n is (w n(1) ,w n(2) …w n(i) …w n(L) ), w n(i) is the delay weight vector W n The delay weight vector coefficient of ; Step 32, determining the error calculation function at time n according to the delay estimation model function: Among them, e n is the error data at time n, y n is the actuator feedback data collected at time n; Step 33, based on the minimum mean square error MMSE criterion, the minimization objective function at time n is determined according to the error calculation function: Among them, J n (W) is the minimization objective function with the delay weight vector W as the independent variable; Step 34, determining the weight iteration function for weight vector iteration at time n according to the minimization objective function: Among them, μ n is the step length data at time n; Step 35, setting the step length in the weight iteration function to the normalized step length, and determining the step length calculation function to be: Among them, α and β are preset constants, 0<α<2, 0≤β, x n Send data to the actuator collected at time n.
4. The delay estimation method of the automatic driving vehicle drive-by-wire system according to claim 1, characterized in that: The first vector S t Substitute the predetermined step length calculation function to calculate the corresponding first step length data sp t , specifically including: Assume preset constants α=1, β=0.00001; For the first vector S t Transpose to get the corresponding first transposed vector The preset constants α and β, the first transposed vector The first vector S t The first sent data s t , substitute into the step length calculation function Get the corresponding first step length data sp n for:
5. The method for estimating delay of an automatic driving vehicle drive-by-wire system according to claim 1, characterized in that: The first vector S t and the first feedback data f t Substitute the predetermined error calculation function to calculate the corresponding first error data err t , specifically including: Get the delay weight vector B obtained at the previous time n-1 t ; For the delay weight vector B t Transpose to get the corresponding delay weight transpose vector Transpose the delay weight vector The first vector S t and the first feedback data f t Substitute into the error calculation function Get the corresponding first error data err t for:
6. The delay estimation method of the automatic driving vehicle drive-by-wire system according to claim 1, characterized in that: The first vector S t , the first step length data sp t and the first error data err t Substitute the predetermined weight iteration function to calculate the latest first delay weight vector B t+1 , specifically including: Get the delay weight vector B obtained at the previous time n-1 t ; And the delay weight vector B t , the first vector S t , the first step length data sp t and the first error data err t Substitute into the weight iteration function W n+1 =W n +μ n e n X n Get the corresponding first delay weight vector B t+1 For: B t+1 =B t +sp t err t S t .
7. The method for estimating delay of an automatic driving vehicle drive-by-wire system according to claim 1, characterized in that: The method further comprises: A plurality of delay estimation data sets R stored according to the executor identification data pairs t Perform a query and draw the delay curve of the corresponding actuator under different working conditions based on the query results.
8. The method for estimating delay of an automatic driving vehicle drive-by-wire system according to claim 1, characterized in that: The method further comprises: A plurality of delay estimation data groups R stored according to actuator identification data and operating condition data t Perform a query and draw the delay curve of the corresponding actuator under the specified working conditions based on the query results.
9. An electronic device, characterized in that: include: memory, processors, and transceivers; The processor is used to couple with the memory, read and execute instructions in the memory, so as to implement the method steps described in any one of claims 1 to 8; The transceiver is coupled to the processor, and the processor controls the transceiver to send and receive messages.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, which, when executed by a computer, enable the computer to execute the method according to any one of claims 1 to 8.
