Speed Control Method and Device for Driverless Vehicle

Abstract

The present application discloses a speed control method and device for an autonomous vehicle, relating to the field of autonomous driving technology. The method includes: obtaining the current desired acceleration and the current actual acceleration of the autonomous vehicle at the current moment; obtaining the target tracking acceleration data of the autonomous vehicle at the next moment based on the current desired acceleration; obtaining the target observed acceleration data and the target disturbance data of the autonomous vehicle at the next moment based on the current actual acceleration; determining the target tracking error data of the autonomous vehicle at the next moment by using the target tracking acceleration data and the target observed acceleration data; and determining the target speed control parameters of the autonomous vehicle based on the target tracking error data and the target disturbance data. This solution effectively reduces the speed control deviation caused by errors and disturbances during the speed control process of the autonomous vehicle.

Description

Technical Field

[0001] This application relates to the field of driverless technologies, and in particular, to a speed control method and device for a driverless vehicle. Background Art

[0002] The operating road surfaces of driverless mining vehicles are relatively poor, with potholed scenarios and many slopes. In addition, the wire control consistency of mining vehicles is relatively poor, and there are certain performance deviations in braking and driving between different vehicles. Using a conventional PID controller cannot well identify the external disturbances caused by such external environments and the internal disturbances caused by the relatively poor wire control consistency of the vehicle itself, resulting in tracking speed deviations due to these disturbances during the actual vehicle speed following control process, or frequent alternating control of the actuator. Summary of the Invention

[0003] In order to overcome the deficiencies of the prior art, this application provides a speed control method and device for a driverless vehicle.

[0004] The technical solution adopted by this application to solve its technical problems is as follows:

[0005] In a first aspect, a speed control method for a driverless vehicle is provided, including:

[0006] Obtain the current desired acceleration and the current actual acceleration of the driverless vehicle at the current moment;

[0007] Based on the current desired acceleration, obtain the target tracking acceleration data of the driverless vehicle at the next moment;

[0008] Based on the current actual acceleration, obtain the target observed acceleration data of the driverless vehicle at the next moment and the target disturbance data at the next moment;

[0009] Use the target tracking acceleration data and the target observed acceleration data to determine the target tracking error data of the driverless vehicle at the next moment;

[0010] Based on the target tracking error data and the target disturbance data, determine the target speed control parameter of the driverless vehicle.

[0011] Further, obtaining the target tracking acceleration data of the driverless vehicle at the next moment includes: obtaining the target tracking acceleration data of the driverless vehicle at the next moment based on an objective function, where the objective function is used to adjust the differential operation parameter values, and the differential operation parameter values include a speed coefficient and a filtering factor. The speed coefficient is related to the tracking speed, and the filtering factor is related to the smoothness of the tracking.

[0012] Further, obtaining the target observed acceleration data and the target disturbance data of the driverless vehicle at the next moment includes: obtaining the target observed acceleration data and the target disturbance data of the driverless vehicle at the next moment based on a correction function, where the correction function is used to adjust the gain according to the tracking error, and the tracking error and the gain are inversely correlated.

[0013] Further, the target tracking acceleration data includes: target tracking acceleration and target tracking jerk; and / or, the target observed acceleration data includes: target observed acceleration and target observed jerk; and / or, the target tracking error data includes: target tracking acceleration error and target tracking jerk error.

[0014] Further, the target tracking acceleration data includes target tracking acceleration and target tracking jerk. Obtaining the target tracking acceleration data of the driverless vehicle at the next moment includes: obtaining the target tracking acceleration of the driverless vehicle at the next moment by using the current tracking acceleration and the current tracking jerk, where the current tracking acceleration and the current tracking jerk are the target tracking acceleration data corresponding to the current moment of the driverless vehicle; determining the target tracking jerk of the driverless vehicle at the next moment by using the current desired acceleration, the current tracking acceleration, the current tracking jerk, and the target function.

[0015] Further, the target observed acceleration data includes target observed acceleration and target observed jerk. Obtaining the target observed acceleration data and the target disturbance data of the driverless vehicle at the next moment includes: determining the current tracking error of the driverless vehicle at the current moment by using the current observed acceleration and the current actual acceleration, where the current observed acceleration is the observed acceleration of the driverless vehicle at the current moment; obtaining the target observed acceleration of the driverless vehicle at the next moment by using the current observed acceleration, the current observed jerk, and the current tracking error; obtaining the target observed jerk of the driverless vehicle at the next moment according to the current observed jerk, the current tracking error, the current disturbance data, and the correction function, where the current disturbance data is the disturbance data of the driverless vehicle at the current moment; determining the target disturbance data of the driverless vehicle at the next moment by using the current disturbance data, the current tracking error, and the correction function.

[0016] Further, the target tracking error data includes target tracking acceleration error and target tracking jerk error. Based on the target tracking error data and the target disturbance data, determining the target speed control parameter of the driverless vehicle includes: determining the target tracking acceleration error by using the target tracking acceleration and the target observed acceleration; determining the target tracking jerk error by using the target tracking jerk and the target observed jerk; determining the target dynamics model parameter based on the target tracking acceleration error and the target tracking jerk error, and using the correction function; and determining the target speed control parameter based on the target dynamics model parameter and the target disturbance data.

[0017] Further, the target tracking acceleration data of the driverless vehicle at the next moment is obtained through the following formula:

[0018] , where is the current desired acceleration at the current moment, is the tracking acceleration at the current moment, is the tracking jerk at the current moment, is the speed coefficient, is the filtering factor, () is the objective function, is the target tracking acceleration at the next moment, is the target tracking jerk at the next moment.

[0019] Further, the objective function () is characterized in the following manner:

[0020] ,

[0021] where , , d = , = d , y = + , = , = , = , is the filtering factor, is the current desired acceleration at the current moment, is the tracking acceleration at the current moment, is the tracking jerk at the current moment, is the speed coefficient.

[0022] Further, the target observed acceleration data and the target disturbance data of the driverless vehicle at the next moment are obtained through the following formula:

[0023] ;

[0024] ;

[0025] ;

[0026] ;

[0027] where z1(k) is the current observed acceleration at the current moment, z2(k) is the current observed jerk at the current moment, z3(k) is the current disturbance data at the current moment, e(k) is the current tracking error at the current moment, u(k) is the dynamic model parameter at the current moment, is an adjustable weight parameter, is the calibrated system parameter, y(k) is the actual acceleration at the current moment, is the target observed acceleration at the next moment, is the target observed jerk at the next moment, is the target disturbance data at the next moment, () is a correction function, is a constant.

[0028] Further, the correction function () is characterized by the following:

[0029] , = , = , is a constant.

[0030] Further, based on the target tracking error data and the target disturbance data, determining the target speed control parameter of the driverless vehicle includes: using a non - linear feedback model to determine the target speed control parameter of the driverless vehicle based on the target tracking error data and the target disturbance data.

[0031] Further, the non - linear feedback model is characterized by the following:

[0032] ;

[0033] ;

[0034] ;

[0035] ;

[0036] Wherein, is the target tracking acceleration at the next moment, is the target tracking jerk at the next moment, is the target observed acceleration at the next moment, is the target observed jerk at the next moment, is the target disturbance data at the next moment, is the target tracking acceleration error at the next moment, is the target tracking jerk error at the next moment, , , are the system parameter calibration quantities, is a constant, is the target speed control parameter.

[0037] Furthermore, one or more of the following parameters are related to the load state of the driverless vehicle: , , , , , , .

[0038] Furthermore, it further includes: determining whether the current desired acceleration is less than a preset speed value, and / or determining whether the target distance of the driverless vehicle is a short walking distance; if the current desired acceleration is less than the preset speed value, and / or the target distance of the driverless vehicle is a short walking distance, then perform dynamic compensation on the target speed control parameter.

[0039] Furthermore, performing dynamic compensation on the target speed control parameter includes: calculating a dynamic compensation value through the following formula: , k is a calibration quantity, t is a time constant, is the target tracking acceleration error of the driverless vehicle at the next moment; through the following formula, use the dynamic compensation value to perform dynamic compensation on the target speed control parameter: , is the dynamic compensation value, is the target speed control parameter at the next moment, is the compensated target speed control parameter at the next moment.

[0040] In a second aspect, there is provided a speed control device for a driverless vehicle, including:

[0041] A current vehicle data acquisition module, configured to acquire the current desired acceleration and the current actual acceleration of the driverless vehicle at the current moment;

[0042] A desired acceleration data acquisition module, configured to acquire target tracking acceleration data of the driverless vehicle at the next moment based on the current desired acceleration;

[0043] An actual acceleration data acquisition module, configured to acquire target observation acceleration data of the driverless vehicle at the next moment and target disturbance data at the next moment based on the current actual acceleration;

[0044] An error calculation module, configured to determine target tracking error data of the driverless vehicle at the next moment by using the target tracking acceleration data and the target observation acceleration data;

[0045] An output data determination module, configured to determine a target speed control parameter of the driverless vehicle based on the target tracking error data and the target disturbance data.

[0046] In a third aspect, a driverless vehicle is provided, including:

[0047] At least one processor and at least one memory;

[0048] The memory stores executable instructions of the processor;

[0049] The processor is configured to execute the speed control method of the driverless vehicle according to any one of the above.

[0050] The technical solution of the present application provides a speed control method and device for a driverless vehicle. By acquiring the current desired acceleration and the current actual acceleration of the driverless vehicle at the current moment, acquiring target tracking acceleration data of the driverless vehicle at the next moment based on the current desired acceleration, acquiring target observation acceleration data of the driverless vehicle at the next moment and target disturbance data at the next moment based on the current actual acceleration, determining target tracking error data of the driverless vehicle at the next moment by using the target tracking acceleration data and the target observation acceleration data, and determining a target speed control parameter of the driverless vehicle based on the target tracking error data and the target disturbance data, the speed control deviation caused by errors and disturbances is effectively reduced during the speed control process of the driverless vehicle. Description of the Drawings

[0051] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0052] Figure 1 is a schematic flowchart of a speed control method for an autonomous vehicle provided by an embodiment of the present application;

[0053] Figure 2 is a schematic flowchart of another speed control method for an autonomous vehicle provided by an embodiment of the present application;

[0054] Figure 3 is a schematic functional structure diagram of a speed control device for an autonomous vehicle provided by an embodiment of the present application;

[0055] Figure 4 is a schematic functional structure diagram of another speed control device for an autonomous vehicle provided by an embodiment of the present application;

[0056] Figure 5 is a schematic functional structure diagram of an autonomous vehicle provided by an embodiment of the present application. Detailed implementation manners

[0057] To make the objectives, technical solutions, and advantages of the present application clearer, the following will describe the technical solutions of the present application in detail with reference to the drawings and embodiments. Obviously, the described embodiments are only some embodiments of the present application, rather than all embodiments. Based on the embodiments in the present application, all other implementation manners obtained by those of ordinary skill in the art without creative efforts belong to the scope protected by the present application.

[0058] To solve the foregoing problems, referring to Figure 1 , an embodiment of the present application provides a speed control method for an autonomous vehicle, including:

[0059] 101. Obtain the current desired acceleration and the current actual acceleration of the autonomous vehicle at the current moment.

[0060] 102. Based on the current desired acceleration, obtain the target tracking acceleration data of the autonomous vehicle at the next moment.

[0061] 103. Based on the current actual acceleration, obtain the target observed acceleration data of the autonomous vehicle at the next moment and the target disturbance data at the next moment.

[0062] 104. Determine the target tracking error data of the driverless vehicle at the next moment by using the target tracking acceleration data and the target observed acceleration data.

[0063] 105. Determine the target speed control parameter of the driverless vehicle based on the target tracking error data and the target disturbance data.

[0064] The speed control method of the driverless vehicle provided in this embodiment obtains the current desired acceleration and the current actual acceleration of the driverless vehicle at the current moment; based on the current desired acceleration, obtains the target tracking acceleration data of the driverless vehicle at the next moment; based on the current actual acceleration, obtains the target observed acceleration data and the target disturbance data of the driverless vehicle at the next moment; determines the target tracking error data of the driverless vehicle at the next moment by using the target tracking acceleration data and the target observed acceleration data; determines the target speed control parameter of the driverless vehicle based on the target tracking error data and the target disturbance data, starts from the external disturbance of the operating environment and the external disturbance of the vehicle's own by-wire consistency, identifies these disturbances in real time, and compensates for the disturbances, ensuring the smoothness during the automatic driving process.

[0065] As an improvement of the above embodiment, the embodiment of the present invention provides another speed control method for a driverless vehicle. The embodiment of the present invention does not limit the implementation subject of this method, which can be selected by those skilled in the art according to the actual engineering situation. For example, it can be an in-vehicle device, and specifically can be implemented by a driverless management system. Refer to Figure 2 , this method includes:

[0066] Obtain the current desired acceleration and the current actual acceleration of the driverless vehicle at the current moment.

[0067] Optionally, the current desired acceleration can be obtained according to the trajectory of the decision-making plan, and the current actual acceleration can be obtained through inertial navigation.

[0068] 202. Based on the current desired acceleration, obtain the target tracking acceleration data of the driverless vehicle at the next moment.

[0069] In some alternative embodiments, obtaining the target tracking acceleration data of the driverless vehicle at the next moment includes: obtaining the target tracking acceleration data of the driverless vehicle at the next moment based on the objective function, where the objective function is used to adjust the differential operation parameter values, and the differential operation parameter values include a speed coefficient and a filtering factor. The speed coefficient is related to the tracking speed, and the filtering factor is related to the smoothness of the tracking.

[0070] In the embodiments of the present invention, the differential operation parameters are adjusted through an objective function to adapt to different load states of the driverless vehicle. One set of differential operation parameters is used in the light load state, and another set of differential operation parameters is used in the heavy load state, making the control process more accurate and flexible, and greatly improving the speed control efficiency.

[0071] In some alternative embodiments, the target tracking acceleration data includes: target tracking acceleration and target tracking jerk.

[0072] When the target tracking acceleration data includes target tracking acceleration and target tracking jerk, 202 can be implemented (not shown in the figure) through the following process:

[0073] 2021. Obtain the target tracking acceleration of the driverless vehicle at the next moment by using the current tracking acceleration and the current tracking jerk.

[0074] Wherein, the current tracking acceleration and the current tracking jerk are the target tracking acceleration data corresponding to the current moment of the driverless vehicle.

[0075] 2022. Determine the target tracking jerk of the driverless vehicle at the next moment by using the current desired acceleration, the current tracking acceleration, the current tracking jerk, and the objective function.

[0076] In some alternative embodiments, the target tracking acceleration data of the driverless vehicle at the next moment is obtained through the following formula:

[0077] , wherein, is the current desired acceleration at the current moment, is the tracking acceleration at the current moment, is the tracking jerk at the current moment, is the speed coefficient, is the filtering factor, () is the objective function, is the target tracking acceleration at the next moment, is the target tracking jerk at the next moment.

[0078] In some alternative embodiments, the objective function () is characterized by the following:

[0079] ,

[0080] Wherein, ,

[0081] , d = , = d , y = + , = , = , = , is the filtering factor, is the current desired acceleration at the current moment, is the tracking acceleration at the current moment, is the tracking jerk at the current moment, is the velocity coefficient.

[0082] Based on the current actual acceleration, obtain the target observed acceleration data and the target disturbance data of the driverless vehicle at the next moment.

[0083] In some alternative embodiments, obtaining the target observed acceleration data and the target disturbance data of the driverless vehicle at the next moment includes: obtaining the target observed acceleration data and the target disturbance data of the driverless vehicle at the next moment based on a correction function, where the correction function is used to adjust the gain according to the tracking error, and the tracking error and the gain are inversely correlated.

[0084] In the embodiments of the present invention, the differential operation parameters are adjusted through the correction function to adapt to different load states of the driverless vehicle. One set of state observation parameters is used in the light load state, and another set of state observation parameters is used in the heavy load state, making the control process more accurate and flexible, and greatly improving the speed control efficiency.

[0085] In some alternative embodiments, the target observed acceleration data includes: the target observed acceleration and the target observed jerk.

[0086] When the target observed acceleration data includes the target observed acceleration and the target observed jerk, 203 can be implemented but is not limited to the following process (not shown in the figure):

[0087] 2031. Determine the current tracking error of the driverless vehicle at the current moment by using the current observed acceleration and the current actual acceleration.

[0088] Wherein, the current observed acceleration is the observed acceleration of the driverless vehicle at the current moment.

[0089] 2032. Obtain the target observed acceleration of the driverless vehicle at the next moment by using the current observed acceleration, the current observed jerk, and the current tracking error.

[0090] 2033. Obtain the target observed jerk of the driverless vehicle at the next moment according to the current observed jerk, the current tracking error, the current disturbance data, and the correction function.

[0091] Among them, the current disturbance data is the disturbance data of the driverless vehicle at the current moment.

[0092] 2034. Determine the target disturbance data of the driverless vehicle at the next moment by using the current disturbance data, the current tracking error, and the correction function.

[0093] In some alternative embodiments, the target observed acceleration data and the target disturbance data of the driverless vehicle at the next moment are obtained by the following formula:

[0094] ;

[0095] ;

[0096] ;

[0097] ;

[0098] Among them, z1(k) is the current observed acceleration at the current moment, z2(k) is the current observed jerk at the current moment, z3(k) is the current disturbance data at the current moment, e(k) is the current tracking error at the current moment, u(k) is the dynamic model parameter at the current moment, is an adjustable weight parameter, is the calibrated system parameter quantity, y(k) is the actual acceleration at the current moment, is the target observed acceleration at the next moment, is the target observed jerk at the next moment, is the target disturbance data at the next moment, () is the correction function, is a constant.

[0099] In some alternative embodiments, the correction function () is characterized in the following manner:

[0100] , = , = , is a constant.

[0101] Determine the target tracking error data of the driverless vehicle at the next moment by using the target tracking acceleration data and the target observed acceleration data.

[0102] In some alternative embodiments, the target tracking error data includes: the target tracking acceleration error and the target tracking jerk error.

[0103] Determine the target speed control parameters of the driverless vehicle based on the target tracking error data and the target disturbance data.

[0104] When the target tracking error data includes the target tracking acceleration error and the target tracking jerk error, 205 can be implemented but not limited to through the following process (not shown in the figure):

[0105] 2051. Determine the target tracking acceleration error by using the target tracking acceleration and the target observed acceleration.

[0106] 2052. Determine the target tracking jerk error by using the target tracking jerk and the target observed jerk.

[0107] 2053. Based on the target tracking acceleration error and the target tracking jerk error, and use the correction function to determine the target dynamics model parameters.

[0108] 2054. Determine the target speed control parameters based on the target dynamics model parameters and the target disturbance data.

[0109] Specifically, in some alternative embodiments, based on the target tracking error data and the target disturbance data, use the nonlinear feedback model to determine the target speed control parameters of the driverless vehicle.

[0110] In some alternative embodiments, the nonlinear feedback model is characterized in the following manner:

[0111] ;

[0112] ;

[0113] ;

[0114] ;

[0115] Wherein, is the target tracking acceleration at the next moment, is the target tracking jerk at the next moment, is the target observed acceleration at the next moment, is the target observed jerk at the next moment, is the target disturbance data at the next moment, is the target tracking acceleration error at the next moment, is the target tracking jerk error at the next moment, , , are the system parameter calibration quantities, is a constant, is the target speed control parameter.

[0116] In the related art, in order to achieve a relatively good control effect within a relatively wide speed range in PID control, it is necessary to design different basic PID parameters corresponding to different speed segments. A large number of parameters will result in a relatively cumbersome calibration process and require a relatively long calibration time. To solve this problem, in the embodiments of the present invention, one or more of the following parameters are related to the load state of the driverless vehicle: , , , , , , . In a specific embodiment, those skilled in the art adjust these parameters to achieve the adjustment of the objective function and the correction function, and further achieve the adjustment of the PID control parameters, that is, the adjustment of the vehicle speed control parameters.

[0117] In some alternative embodiments, is the adjustable weight standard value of 1 times, is the adjustable weight standard value of 3 times, is one-ninth of the adjustable weight standard value.

[0118] In some alternative embodiments, is the controller bandwidth of 1 time, is the controller bandwidth of 3 times.

[0119] In some alternative embodiments, the filtering factor = 0.02, and the speed coefficient = 0.5.

[0120] Further optionally, the adjustable weight standard value of the driverless vehicle when it is heavily loaded is 0.52, the controller bandwidth is 0.18, = -0.50.

[0121] In some alternative embodiments, the adjustable weight standard value of the driverless vehicle when it is lightly loaded is 0.45, the controller bandwidth is 0.15, = -0.45.

[0122] In some alternative embodiments, = 0.25.

[0123] In this embodiment, the objective function and the correction function are used to perform parameter matching for the state where the mass difference between the empty load and the heavy load of the mining vehicle is very large. One set of basic parameters is used for the empty load, and another set of basic parameters is used for the heavy load, avoiding the coupling of the empty load and the heavy load during the parameter adjustment process, reducing the complexity of parameter adjustment, and reducing the time spent on parameter adjustment.

[0124] Determine whether the current desired acceleration is less than a preset speed value, and / or determine whether the target distance of the driverless vehicle is a short travel distance.

[0125] 207. If the current desired acceleration is less than the preset speed value, and / or the target distance of the driverless vehicle is a short travel distance, then perform dynamic compensation on the target speed control parameter.

[0126] In some alternative embodiments, the dynamic compensation for the target speed control parameter can be, but is not limited to, performed through the following process (not shown in the figure):

[0127] 2071. Calculate the dynamic compensation value through the following formula: , where k is a calibrated quantity, t is a time constant, is the target tracking acceleration error of the driverless vehicle at the next moment.

[0128] 2072. Use the dynamic compensation value to perform dynamic compensation on the target speed control parameter through the following formula: , is the dynamic compensation value, is the target speed control parameter at the next moment, is the compensated target speed control parameter at the next moment.

[0129] When the vehicle starts at a low speed, this embodiment adds recognition and dynamic gain compensation for the problem that the throttle is relatively slow for small target vehicle speeds in the wheel blocking scenario, ensuring a quick start in this scenario.

[0130] The speed control method of the driverless vehicle provided in this embodiment starts from the external disturbances of the operating environment and the external disturbances of the vehicle's own by-wire consistency, real-time identifies these disturbances, and compensates for the disturbances, ensuring the smoothness during the autonomous driving process. Starting from the actual operating state of the driverless mining vehicle, since the mass deviation between the empty and loaded states of the driverless mining vehicle is relatively large, using a set of basic parameters has too high a coupling degree, and there is a problem that the empty and loaded states need to be re-adapted after parameter adjustment. This method decouples the empty and loaded basic parameters, reduces the adaptation difficulty and the adaptation time by changing the control parameters in the empty and loaded states. Real-time detect and identify the external disturbances on the road and the external disturbances caused by the inconsistency of the vehicle's own by-wire.

[0131] Another specific implementation manner of the speed control method of the driverless vehicle provided by the embodiment of the present invention is as follows:

[0132] The first step: Obtain the desired acceleration according to the trajectory of the decision-making plan, and obtain the actual acceleration according to the inertial navigation.

[0133] This desired acceleration is the current desired acceleration, and this actual acceleration is the current actual acceleration.

[0134] Step 2: Extract the differential term corresponding to the desired acceleration to track the jerk, and at the same time obtain the over-processed tracking acceleration. After being processed, this tracking acceleration will be relatively smooth, avoiding large fluctuations in the actuator caused by sudden acceleration and deceleration trajectories.

[0135] Optionally, the process is as follows:

[0136] ,

[0137] where, is the desired acceleration output by decision-making and planning, a1(k) and a2(k) are the tracking acceleration and the tracking jerk respectively, is the velocity coefficient, is the filtering factor.

[0138] fst() is a non-linear function. By adjusting the two parameters of the velocity coefficient and the filtering factor of the fst() function, a fast and smooth tracking acceleration and tracking jerk can be obtained, correcting the sudden acceleration and deceleration trajectory. The structure of the fst() function is as follows:

[0139] , where d = r , = d , y = x1 + , ,

[0140] .

[0141] This differential term tracking jerk is the target tracking acceleration data of the driverless vehicle at the next moment.

[0142] It should be noted that the sign function is a function used to judge the positive and negative of a numerical value. Its basic function is to return the corresponding sign value according to the input numerical value. When the input numerical value is positive, the sign function returns 1. When the input numerical value is zero, it returns 0. When the input numerical value is negative, it returns -1.

[0143] Step 3: Estimate the state of the driverless mining truck and the observed disturbance, the observed acceleration and the observed jerk obtained after the disturbance effect.

[0144] ;

[0145] ;

[0146] ;

[0147] ;

[0148] Among them, , , are the observed acceleration, observed jerk, and observed disturbance respectively. The initial values are generally all 0. e(k) is the tracking error, and u(k) is the system control variable. are three adjustable weight parameters. These three parameters are related to the observer bandwidth. For the convenience of calibration, they are integrated into the bandwidth linear relationship parameter. b0 is the system parameter calibration quantity, y(k) is the actual acceleration, and fal() is a nonlinear function. The structure of fal() is as follows:

[0149] .

[0150] In the embodiment of the present invention, the observed acceleration is the actual acceleration obtained after processing the actual acceleration and taking into account the tracking error.

[0151] Step 4: Obtain the acceleration error based on the tracking acceleration and the observed acceleration, obtain the jerk error by using the tracking jerk and the observed jerk, and use the nonlinear feedback link to obtain the corresponding output quantity. Subtract the product of the observed disturbance and the reciprocal of the system gain from the output quantity to obtain the final system control variable.

[0152] The nonlinear feedback link is expressed as follows:

[0153] ;

[0154] ;

[0155] ;

[0156] ;

[0157] Among them, , is the tracking acceleration and tracking jerk calculated in the second step, is the observed acceleration, observed jerk, and observed disturbance calculated in the third step, is the calibration quantity, is a constant, is the output (a variable of the dynamic model, such as the rolling resistance coefficient).

[0158] For driverless mining trucks, the difference in weight between empty and loaded states is significant. If a single set of parameters is used, it cannot meet the requirements of precise control. Moreover, using the same set of parameters for empty and loaded states will increase the difficulty and complexity of calibration. After calibrating the loaded state and modifying the parameters during the control calibration, since these parameters also affect the loaded state, it is necessary to re-verify the loaded scenario again.

[0159] In the scenario of following a small desired speed, in related technologies, the accelerator is slowly depressed according to the small desired speed. At this time, if the wheels are blocked by potholes or stones, the vehicle cannot start with a small throttle, and starting with a slow throttle will be relatively slow, causing bottlenecks in operations. For example, for the issued acceleration trajectory (the desired speed to be reached is less than the speed threshold), but the actual speed is less than the speed threshold by 0.3. To identify this scenario, in this embodiment, a gain compensation is performed on the acceleration deviation parameter term. After triggering this scenario, the gain compensation will increase dynamically according to time, and the dynamic gain will be recovered after the vehicle starts, so as to ensure the subsequent control accuracy.

[0160] The dynamic compensation formula is as follows:

[0161] , where k is a calibrated quantity, t is a time constant, is the deviation between the tracking acceleration and the observed acceleration calculated in the fourth step.

[0162] As described above, using the speed control method for driverless vehicles provided by the embodiments of the present invention can more quickly identify external disturbances in the operating environment of driverless mining trucks and internal disturbances of the mining trucks themselves, and compensate for these disturbances to achieve the purpose of precise control. According to the special phenomenon that the mass difference between the empty and loaded states of on-site operating vehicles is very large, a function with adjustable parameters is designed to facilitate setting two sets of basic parameters for empty and loaded states, decoupling the calibration of empty and loaded states, and reducing the calibration workload. According to the individual special business scenarios of driverless mining trucks, such as small desired speeds and short travel distances, the problem that the accelerator is depressed too slowly in related technology algorithms when the wheels are blocked is identified, and dynamic compensation is performed on this scenario, increasing the gain compensation ratio according to the time after receiving the desired speed.

[0163] To cooperate with the implementation of the above speed control method for driverless vehicles, an embodiment of the present invention provides a speed control device for driverless vehicles. Refer to Figure 3 , this device includes:

[0164] A current vehicle data acquisition module 31, configured to acquire the current desired acceleration and the current actual acceleration of the driverless vehicle at the current moment.

[0165] A desired acceleration data acquisition module 32, configured to acquire the target tracking acceleration data of the driverless vehicle at the next moment based on the current desired acceleration.

[0166] The actual acceleration data acquisition module 33 is configured to obtain the target observation acceleration data and the target disturbance data of the driverless vehicle at the next moment based on the current actual acceleration.

[0167] The error calculation module 34 is configured to determine the target tracking error data of the driverless vehicle at the next moment by using the target tracking acceleration data and the target observation acceleration data.

[0168] The output data determination module 35 is configured to determine the target speed control parameter of the driverless vehicle based on the target tracking error data and the target disturbance data.

[0169] For the speed control device of the driverless vehicle provided in this embodiment, the current vehicle data acquisition module acquires the current desired acceleration and the current actual acceleration of the driverless vehicle at the current moment. The desired acceleration data acquisition module acquires the target tracking acceleration data of the driverless vehicle at the next moment based on the current desired acceleration. The actual acceleration data acquisition module acquires the target observation acceleration data and the target disturbance data of the driverless vehicle at the next moment based on the current actual acceleration. The error calculation module determines the target tracking error data of the driverless vehicle at the next moment by using the target tracking acceleration data and the target observation acceleration data. The output data determination module determines the target speed control parameter of the driverless vehicle based on the target tracking error data and the target disturbance data. This device starts from the external disturbances of the operating environment and the external disturbances of the vehicle's own by-wire consistency, real-time identifies these disturbances, and compensates for the disturbances, ensuring the smoothness during the automatic driving process.

[0170] As an improvement of the above embodiment, the embodiment of the present invention provides another speed control device for a driverless vehicle. Refer to Figure 4 , and this device includes:

[0171] The current vehicle data acquisition module 41 is configured to acquire the current desired acceleration and the current actual acceleration of the driverless vehicle at the current moment.

[0172] The desired acceleration data acquisition module 42 is configured to acquire the target tracking acceleration data of the driverless vehicle at the next moment based on the current desired acceleration.

[0173] In some optional embodiments, acquiring the target tracking acceleration data of the driverless vehicle at the next moment is: based on the objective function, using differential operation to acquire the target tracking acceleration data of the driverless vehicle at the next moment. The objective function is used to adjust the differential operation parameter values. The differential operation parameter values include a speed coefficient and a filtering factor. The speed coefficient is related to the tracking speed, and the filtering factor is related to the smoothness of the tracking.

[0174] In some alternative embodiments, the target tracking acceleration data includes target tracking acceleration and target tracking jerk, and the desired acceleration data acquisition module 42 includes:

[0175] A target tracking acceleration acquisition sub-module 421, configured to obtain the target tracking acceleration of the driverless vehicle at the next moment by using the current tracking acceleration and the current tracking jerk, where the current tracking acceleration and the current tracking jerk are the target tracking acceleration data corresponding to the current moment of the driverless vehicle.

[0176] A target tracking acceleration determination sub-module 422, configured to determine the target tracking jerk of the driverless vehicle at the next moment by using the current desired acceleration, the current tracking acceleration, the current tracking jerk, and the objective function.

[0177] In some alternative embodiments, the target tracking acceleration data of the driverless vehicle at the next moment is obtained by the following formula:

[0178] , where is the current desired acceleration at the current moment, is the tracking acceleration at the current moment, is the tracking jerk at the current moment, is the speed coefficient, is the filtering factor, () is the objective function, is the target tracking acceleration at the next moment, is the target tracking jerk at the next moment.

[0179] Among them, the objective function () is characterized by the following:

[0180] , where

[0181] , , d = , = d , y = + , = , = , = , is the filtering factor, is the current desired acceleration at the current moment, is the tracking acceleration at the current moment, is the tracking jerk at the current moment, is the speed coefficient.

[0182] The actual acceleration data acquisition module 43 is used to obtain the target observed acceleration data and the target disturbance data of the driverless vehicle at the next moment based on the current actual acceleration.

[0183] In some alternative embodiments, obtaining the target observed acceleration data and the target disturbance data of the driverless vehicle at the next moment is as follows: obtaining the target observed acceleration data and the target disturbance data of the driverless vehicle at the next moment based on a correction function, where the correction function is used to adjust the gain according to the tracking error, and the tracking error and the gain are inversely correlated.

[0184] In some alternative embodiments, the target observed acceleration data includes the target observed acceleration and the target observed jerk, and the actual acceleration data acquisition module 43 includes:

[0185] The tracking error determination sub-module 431 is used to determine the current tracking error of the driverless vehicle at the current moment by using the current observed acceleration and the current actual acceleration, where the current observed acceleration is the observed acceleration of the driverless vehicle at the current moment.

[0186] The target observed acceleration acquisition sub-module 432 is used to obtain the target observed acceleration of the driverless vehicle at the next moment by using the current observed acceleration, the current observed jerk, and the current tracking error.

[0187] The target observed jerk acquisition sub-module 433 is used to obtain the target observed jerk of the driverless vehicle at the next moment according to the current observed jerk, the current tracking error, the current disturbance data, and the correction function, where the current disturbance data is the disturbance data of the driverless vehicle at the current moment.

[0188] The target disturbance data determination sub-module 434 is used to determine the target disturbance data of the driverless vehicle at the next moment by using the current disturbance data, the current tracking error, and the correction function.

[0189] In some alternative embodiments, the target observed acceleration data and the target disturbance data of the driverless vehicle at the next moment are obtained through the following formulas:

[0190] ;

[0191] ;

[0192] ;

[0193] ;

[0194] Where, z1(k) is the current observed acceleration at the current moment, z2(k) is the current observed jerk at the current moment, z3(k) is the current disturbance data at the current moment, e(k) is the current tracking error at the current moment, and u(k) is the dynamic model parameter at the current moment. is an adjustable weight parameter. is the calibrated system parameter, and y(k) is the actual acceleration at the current moment. is the target observed acceleration at the next moment. is the target observed jerk at the next moment. is the target disturbance data at the next moment. () is a correction function. is a constant.

[0195] Among them, the correction function () is characterized in the following way:

[0196] , = , = , is a constant.

[0197] The error calculation module 44 is used to determine the target tracking error data of the driverless vehicle at the next moment by using the target tracking acceleration data and the target observed acceleration data.

[0198] In some alternative embodiments, the target tracking acceleration data includes: the target tracking acceleration and the target tracking jerk. And / or,

[0199] The target observed acceleration data includes: the target observed acceleration and the target observed jerk. And / or,

[0200] The target tracking error data includes: the target tracking acceleration error and the target tracking jerk error.

[0201] The output data determination module 45 is used to determine the target speed control parameter of the driverless vehicle based on the target tracking error data and the target disturbance data.

[0202] In some alternative embodiments, the target tracking error data includes the target tracking acceleration error and the target tracking jerk error, and the output data determination module 45 includes:

[0203] The target tracking acceleration error determination sub-module 451 is used to determine the target tracking acceleration error by using the target tracking acceleration and the target observed acceleration.

[0204] The jerk error determination sub-module 452 for target tracking is configured to determine the jerk error of target tracking by using the jerk of target tracking and the jerk of target observation.

[0205] The target dynamics model parameter determination sub-module 453 is configured to determine the target dynamics model parameters based on the acceleration error of target tracking and the jerk error of target tracking, and by using a correction function.

[0206] The target speed control parameter determination sub-module 454 is configured to determine the target speed control parameters based on the target dynamics model parameters and the target disturbance data.

[0207] Specifically, in some alternative embodiments, the output data determination module 45 determines the target speed control parameters of the driverless vehicle by using a non-linear feedback model based on the target tracking error data and the target disturbance data.

[0208] In some alternative embodiments, the non-linear feedback model is characterized in the following manner:

[0209] ;

[0210] ;

[0211] ;

[0212] ;

[0213] where, is the target tracking acceleration at the next moment, is the jerk of target tracking at the next moment, is the target observation acceleration at the next moment, is the jerk of target observation at the next moment, is the target disturbance data at the next moment, is the acceleration error of target tracking at the next moment, is the jerk error of target tracking at the next moment, 、 、 are the calibrated system parameter values, is a constant, is the target speed control parameter.

[0214] The compensation module 46 is configured to determine whether the current desired acceleration is less than a preset speed value, and / or determine whether the target distance of the driverless vehicle is a short travel distance. If the current desired acceleration is less than the preset speed value, and / or the target distance of the driverless vehicle is a short travel distance, then dynamic compensation is performed on the target speed control parameters.

[0215] In some alternative embodiments, the compensation module 46 includes:

[0216] A compensation value calculation sub-module 461, configured to calculate a dynamic compensation value through the following formula: , where k is a calibrated quantity, t is a time constant, is the target tracking acceleration error of the driverless vehicle at the next moment.

[0217] A compensation sub-module 462, configured to dynamically compensate the target speed control parameter by using the dynamic compensation value through the following formula: , is the dynamic compensation value, is the target speed control parameter at the next moment, is the compensated target speed control parameter at the next moment.

[0218] It should be noted that one or more of the following parameters are related to the load state of the driverless vehicle: 、 、 、 .

[0219] For the speed control device of the driverless vehicle provided in this embodiment, the current vehicle data acquisition module acquires the current desired acceleration and the current actual acceleration of the driverless vehicle at the current moment. The desired acceleration data acquisition module acquires the target tracking acceleration data of the driverless vehicle at the next moment based on the current desired acceleration. The actual acceleration data acquisition module acquires the target observed acceleration data and the target disturbance data at the next moment of the driverless vehicle based on the current actual acceleration. The error calculation module determines the target tracking error data of the driverless vehicle at the next moment by using the target tracking acceleration data and the target observed acceleration data. The output data determination module determines the target speed control parameter of the driverless vehicle based on the target tracking error data and the target disturbance data. The compensation module compensates the vehicle speed control parameter when the vehicle starts at a low speed. The device starts from the external disturbances of the operating environment and the external disturbances of the vehicle's own wire control consistency, identifies these disturbances in real time, and compensates for the disturbances, ensuring the smoothness of the automatic driving process.

[0220] Based on the same inventive concept, as Figure 5 shown, the present application further provides a driverless vehicle, including:

[0221] At least one processor 51 and at least one memory 52;

[0222] The memory stores executable instructions of the processor;

[0223] The processor is configured to execute the speed control method for the driverless vehicle provided in the above embodiments.

[0224] For the driverless vehicle provided in the embodiments of the present application, the executable instructions of the processor are stored in a memory. When the executable instructions are executed, the processor can start from external disturbances in the operating environment and external disturbances of the vehicle's own by-wire consistency, identify these disturbances in real time, and compensate for the disturbances, ensuring the smoothness during the autonomous driving process.

[0225] It should be noted that in the description of the present application, the terms "first", "second", etc. are only used for descriptive purposes and cannot be construed as indicating or implying relative importance. In addition, in the description of the present application, unless otherwise specified, the meaning of "a plurality of" refers to at least two.

[0226] It can be understood that the same or similar parts in the above embodiments can be referred to each other, and the content not detailed in some embodiments can be seen in the same or similar content in other embodiments.

Claims

1. A speed control method for an autonomous vehicle, characterized in that, Including: Obtaining the current desired acceleration and the current actual acceleration of the driverless vehicle at the current moment; Based on the current desired acceleration, obtaining the target tracking acceleration data of the driverless vehicle at the next moment; Based on the current actual acceleration, obtaining the target observed acceleration data and the target disturbance data of the driverless vehicle at the next moment; Using the target tracking acceleration data and the target observed acceleration data to determine the target tracking error data of the driverless vehicle at the next moment; Based on the target tracking error data and the target disturbance data, determining the target speed control parameter of the driverless vehicle; The target tracking acceleration data includes a target tracking acceleration and a target tracking jerk. Obtaining the target tracking acceleration data of the driverless vehicle at the next moment includes: using the current tracking acceleration and the current tracking jerk to obtain the target tracking acceleration of the driverless vehicle at the next moment, where the current tracking acceleration and the current tracking jerk are the target tracking acceleration data corresponding to the current moment of the driverless vehicle; using the current desired acceleration, the current tracking acceleration, the current tracking jerk, and an objective function to determine the target tracking jerk of the driverless vehicle at the next moment.

2. The method according to claim 1, characterized in that, Obtaining the target tracking acceleration data of the driverless vehicle at the next moment includes: Obtaining the target tracking acceleration data of the driverless vehicle at the next moment based on an objective function, where the objective function is used to adjust the differential operation parameter values, and the differential operation parameter values include a speed coefficient and a filtering factor. The speed coefficient is related to the tracking speed, and the filtering factor is related to the smoothness of the tracking.

3. The method according to claim 2, characterized in that Obtaining the target observed acceleration data and the target disturbance data of the driverless vehicle at the next moment includes: Obtaining the target observed acceleration data and the target disturbance data of the driverless vehicle at the next moment based on a correction function, where the correction function is used to perform gain adjustment according to the tracking error, and the tracking error and the gain are inversely related.

4. The method according to claim 1 or 2, wherein the target tracking acceleration data includes: Target tracking acceleration and target tracking jerk; And / or, The target observed acceleration data includes: target observed acceleration and target observed jerk; and / or, The target tracking error data includes: target tracking acceleration error and target tracking jerk error.

5. The method according to claim 3, characterized in that The target observed acceleration data includes target observed acceleration and target observed jerk, Obtaining the target observed acceleration data and the target disturbance data of the driverless vehicle at the next moment includes: Using the current observed acceleration and the current actual acceleration to determine the current tracking error of the driverless vehicle at the current moment, where the current observed acceleration is the observed acceleration of the driverless vehicle at the current moment; Using the current observed acceleration, the current observed jerk, and the current tracking error to obtain the target observed acceleration of the driverless vehicle at the next moment; Obtain the target observed jerk of the driverless vehicle at the next moment according to the current observed jerk, the current tracking error, the current disturbance data, and the correction function, where the current disturbance data is the disturbance data of the driverless vehicle at the current moment; Determine the target disturbance data of the driverless vehicle at the next moment by using the current disturbance data, the current tracking error, and the correction function.

6. The method according to claim 5, characterized in that, The target tracking error data includes a target tracking acceleration error and a target tracking jerk error. Based on the target tracking error data and the target disturbance data, determining the target speed control parameter of the driverless vehicle includes: Determine the target tracking acceleration error by using the target tracking acceleration and the target observed acceleration; Determine the target tracking jerk error by using the target tracking jerk and the target observed jerk; Based on the target tracking acceleration error and the target tracking jerk error, and use the correction function to determine the target dynamic model parameters; Determine the target speed control parameter based on the target dynamic model parameters and the target disturbance data.

7. The method according to claim 6, wherein Obtain the target tracking acceleration data of the driverless vehicle at the next moment through the following formula: , where is the current desired acceleration at the current moment, is the tracking acceleration at the current moment, is the tracking jerk at the current moment, is the speed coefficient, is the filtering factor, () is the objective function, is the target tracking acceleration at the next moment, is the target tracking jerk at the next moment.

8. The method according to claim 7, wherein The objective function is characterized by the following: ， Among them, , , d = , = d , y = + , = , = , = , is the filtering factor, is the current expected acceleration at the current moment, is the tracking acceleration at the current moment, is the tracking jerk at the current moment, is the velocity coefficient.

9. The method according to claim 5, characterized in that Obtain the target observed acceleration data and the target disturbance data at the next moment of the driverless vehicle through the following formula: ； ； ； ； Among them, z1(k) is the current observed acceleration at the current moment, z2(k) is the current observed jerk at the current moment, z3(k) is the current disturbance data at the current moment, e(k) is the current tracking error at the current moment, and u(k) is the dynamic model parameter at the current moment. is the adjustable weight parameter. is the calibrated value of the system parameter, and y(k) is the actual acceleration at the current moment. is the target observed acceleration at the next moment. is the target observed jerk at the next moment. is the target disturbance data at the next moment. () is the correction function. is a constant.

10. The method according to claim 9, characterized in that, The correction function is characterized by the following: , = , = , is a constant.

11. The method according to claim 9, wherein Based on the target tracking error data and the target disturbance data, determining the target speed control parameter of the driverless vehicle includes: Based on the target tracking error data and the target disturbance data, use a nonlinear feedback model to determine the target speed control parameter of the driverless vehicle.

12. The method according to claim 11, characterized in that, The nonlinear feedback model is characterized in the following manner: ； ； ； ； Among them, is the target tracking acceleration at the next moment, is the target tracking jerk at the next moment, is the target observation acceleration at the next moment, is the target observation jerk at the next moment, is the target disturbance data at the next moment, is the target tracking acceleration error at the next moment, is the target tracking jerk error at the next moment, , , are the calibrated system parameter values, is a constant, is the target speed control parameter.

13. The method according to claim 12, characterized in that, One or more of the following parameters are related to the load status of the driverless vehicle: , , , , , , .

14. The method according to claim 12, characterized in that, Further include: Judge whether the current desired acceleration is less than a preset speed value, and / or judge whether the target distance of the driverless vehicle is a short walking distance; If the current desired acceleration is less than the preset speed value, and / or the target distance of the driverless vehicle is a short walking distance, then perform dynamic compensation on the target speed control parameter.

15. The method according to claim 14, wherein Performing dynamic compensation on the target speed control parameter includes: The dynamic compensation value is calculated by the following formula: , where k is the calibrated quantity and t is the time constant, is the target tracking acceleration error of the driverless vehicle at the next moment; The target speed control parameter is dynamically compensated by using the dynamic compensation value through the following formula: , is the dynamic compensation value, is the target speed control parameter at the next moment, is the compensated target speed control parameter at the next moment.

16. A speed control device for an autonomous vehicle, characterized in that, Include: A current vehicle data acquisition module, configured to acquire the current desired acceleration and the current actual acceleration of the driverless vehicle at the current moment; A desired acceleration data acquisition module, configured to acquire the target tracking acceleration data of the driverless vehicle at the next moment based on the current desired acceleration; An actual acceleration data acquisition module, configured to acquire the target observed acceleration data and the target disturbance data at the next moment of the driverless vehicle based on the current actual acceleration; An error calculation module, configured to determine the target tracking error data of the driverless vehicle at the next moment by using the target tracking acceleration data and the target observed acceleration data; An output data determination module, configured to determine the target speed control parameter of the driverless vehicle based on the target tracking error data and the target disturbance data; The target tracking acceleration data includes target tracking acceleration and target tracking jerk. The desired acceleration data acquisition module includes: a target tracking acceleration acquisition sub-module, configured to obtain the target tracking acceleration of the driverless vehicle at the next moment by using the current tracking acceleration and the current tracking jerk, where the current tracking acceleration and the current tracking jerk are the target tracking acceleration data corresponding to the current moment of the driverless vehicle; a target tracking jerk determination sub-module, configured to determine the target tracking jerk of the driverless vehicle at the next moment by using the current desired acceleration, the current tracking acceleration, the current tracking jerk, and an objective function.

Images