A Longitudinal Control Method and System for New Energy Heavy Trucks Based on Disturbance Observation and Feedforward

Abstract

This application relates to the field of vehicle control technology and discloses a longitudinal control method and system for new energy heavy-duty trucks based on disturbance observation and feedforward. The method includes: processing and converting vehicle speed using slip ratio to obtain actual vehicle speed and actual vehicle acceleration; processing the target vehicle speed signal to obtain reference acceleration and reference vehicle speed; inputting the actual vehicle speed, the command from the previous control cycle, and the brake pedal opening into a third-order linear extended state observer model to obtain the equivalent compensation torque for slow-varying disturbances; calculating the feedforward torque and the final feedback torque based on the reference acceleration and speed tracking error, and merging them with the equivalent compensation torque to obtain the original torque command; calculating a comprehensive limiting factor using speed tracking error and acceleration deviation to limit the original torque command and obtain candidate values ​​for the torque command, thereby obtaining the feedback torque command. This invention can separate road resistance and transient braking force, avoid vehicle speed overshoot, and improve control smoothness.

Description

Technical Field

[0001] This invention relates to the field of vehicle control technology, specifically to a longitudinal control method and system for new energy heavy-duty trucks based on disturbance observation and feedforward. Background Technology

[0002] New energy heavy-duty trucks are widely used in logistics transportation. The longitudinal control system of new energy heavy-duty trucks directly affects the smoothness and safety of vehicle driving. Existing longitudinal control of new energy heavy-duty trucks usually relies on conventional proportional-integral-derivative control or a basic feedforward and feedback combined architecture. In actual operation, new energy heavy-duty trucks often face complex road environments. Changes in road slope will introduce slow-varying resistance disturbances. At the same time, new energy heavy-duty trucks generally use pneumatic braking systems. Pneumatic braking systems have asymmetrical time delay characteristics during the braking pressure build-up and depressurization process. Existing state observation methods often find it difficult to effectively separate the transient force generated by pneumatic braking from the slow-varying resistance generated by road slope, resulting in inaccurate calculation of disturbance compensation torque.

[0003] During long-term speed tracking, traditional integral control is prone to accumulating integral saturation, making it difficult for the control system to reduce the control quantity in time when the vehicle speed approaches the target value. Existing torque limiting strategies usually use fixed rate of change boundaries, which cannot dynamically adjust the limit range of torque commands according to the current speed tracking error and acceleration deviation. This causes new energy heavy trucks to easily generate longitudinal control speed overshoot during the speed following phase, thereby reducing the continuity of torque output and overall control stability during vehicle operation.

[0004] Therefore, this invention proposes a longitudinal control method and system for new energy heavy trucks based on disturbance observation and feedforward to address the shortcomings of existing technologies. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a longitudinal control method and system for new energy heavy-duty trucks based on disturbance observation and feedforward. This solves the problems of existing longitudinal control methods for new energy heavy-duty trucks, which have difficulty effectively separating the transient force of air pressure braking from the slow-varying resistance generated by road slope, leading to inaccurate calculation of disturbance compensation torque, and are prone to integral saturation and speed overshoot during speed tracking.

[0006] To achieve the above objectives, the present invention provides the following technical solution: The first aspect of the present invention is a longitudinal control method for new energy heavy-duty trucks based on disturbance observation and feedforward, comprising:

[0007] The system collects the converted vehicle speeds of the driven wheels and the driving wheels, calculates the slip ratio based on these converted speeds, and processes the converted vehicle speeds of the driven wheels and the driving wheels using the slip ratio to obtain the actual vehicle speed and actual vehicle acceleration.

[0008] Receive the target vehicle speed signal, process the target vehicle speed signal, and obtain the reference acceleration and reference vehicle speed;

[0009] The actual vehicle speed, the feedback torque command from the previous control cycle, and the brake pedal opening are input into the third-order linear extended state observer model for calculation to obtain the equivalent compensation torque for slow-varying disturbances.

[0010] The feedforward torque is calculated based on the reference acceleration. The speed error is obtained by calculating the difference between the reference vehicle speed and the actual vehicle speed. The final feedback torque is calculated based on the speed error. The feedforward torque, the final feedback torque and the slow-change disturbance equivalent compensation torque are combined to obtain the original torque command.

[0011] The acceleration deviation is obtained by calculating the difference between the reference acceleration and the actual vehicle acceleration. The comprehensive limiting factor is calculated based on the speed error and the acceleration deviation. The original torque command is limited by the comprehensive limiting factor to obtain the candidate value of the torque command.

[0012] The candidate values ​​of the torque command are processed to obtain the feedback torque command.

[0013] Preferably, the process of processing the converted vehicle speeds of the driven wheels and the driving wheels using the slip ratio to obtain the actual vehicle speed and actual vehicle acceleration specifically includes:

[0014] Compare the slip ratio with a preset slip ratio threshold;

[0015] When the slip ratio is greater than the slip ratio threshold, the preset basic drive wheel weight is attenuated to obtain the actual drive wheel weight;

[0016] When the slip ratio is less than or equal to the slip ratio threshold, the weight of the basic drive wheel is directly assigned to the weight of the actual drive wheel;

[0017] Obtain the preset constant 1, calculate the difference between constant 1 and the actual driving wheel weight, and obtain the actual driven wheel weight;

[0018] The original vehicle speed is obtained by weighting the converted vehicle speed of the driven wheels and the converted vehicle speed of the driving wheels using the actual driven wheel weights and the actual driving wheel weights.

[0019] The original vehicle speed is low-pass filtered to obtain the actual vehicle speed, and the actual vehicle speed is differentiated to obtain the actual vehicle acceleration.

[0020] Preferably, processing the target vehicle speed signal to obtain the reference acceleration and reference vehicle speed specifically includes:

[0021] The original impact intensity is obtained based on the changing trend of the target vehicle speed signal;

[0022] The original impact is constrained according to a preset impact limit value to obtain a reference impact.

[0023] The temporary reference acceleration is obtained by integrating the reference impact, and the temporary reference acceleration is constrained according to a preset acceleration limit value to obtain the reference acceleration.

[0024] The reference acceleration is integrated to obtain a temporary reference speed, and the temporary reference speed is speed-limited according to the set maximum driving speed to obtain the reference speed.

[0025] Preferably, the actual vehicle speed, the feedback torque command from the previous control cycle, and the brake pedal opening are input into a third-order linear extended state observer model for calculation to obtain the equivalent compensation torque for slow-varying disturbances. Specifically, this includes:

[0026] The third-order linear extended state observer model includes vehicle speed state variables, slow-varying lumped disturbance state variables, and transient air braking force equivalent torque state variables.

[0027] Determine the direction of change in the brake pedal opening;

[0028] When the brake pedal opening increases, a preset brake pressure build-up time constant is selected;

[0029] When the brake pedal opening decreases, a preset brake pressure relief time constant is selected;

[0030] The actual vehicle speed, the feedback torque command of the previous control cycle, and the brake pedal opening are used as input items. Combined with the selected brake pressure build-up time constant or brake pressure relief time constant, discretization iterative calculation is performed in the third-order linear extended state observer model based on the forward Euler method. The vehicle speed state variable, the slow-varying lumped disturbance state variable, and the transient air braking force equivalent torque state variable are updated simultaneously.

[0031] Extract the updated estimated values ​​of the slow-varying lumped disturbance state variables, multiply the extracted estimated values ​​of the slow-varying lumped disturbance state variables by the nominal equivalent mass, and obtain the slow-varying disturbance equivalent compensation torque.

[0032] By incorporating asymmetric fluid pressure build-up time parameters, such as the braking pressure build-up time constant and the braking pressure relief time constant, into the model calculation, the road slope resistance and transient braking force can be effectively separated.

[0033] Preferably, the calculation of the final feedback torque based on the speed error specifically includes:

[0034] Calculate the proportional-derivative feedback torque based on the speed error;

[0035] Based on multiple preset speed error thresholds, conditional integration is performed on the speed error to generate an integral accumulation term;

[0036] The final feedback torque is obtained by combining the proportional-derivative feedback torque and the integral accumulation term.

[0037] Preferably, the conditional integration calculation of the speed error based on multiple preset speed error thresholds to generate an integral accumulation term specifically includes:

[0038] Determine the absolute value of the speed error in the current control cycle;

[0039] When the absolute value of the speed error is less than the lower limit of the speed error, the speed error is accumulated into the integral accumulation term with full weight.

[0040] When the absolute value of the speed error is greater than or equal to the lower limit of the speed error and less than the preset upper limit, the speed error multiplied by the attenuation weight is added to the integral accumulation term;

[0041] When the absolute value of the speed error is greater than or equal to the preset upper limit, the integral accumulation term is kept at the current value, and the accumulation of the speed error for the current control cycle is stopped.

[0042] When the product of the speed error of the current control cycle and the speed error of the previous control cycle is less than zero, the integral accumulation term is cleared.

[0043] The long-term velocity error integral is calculated by using piecewise integral weight constraints, and the historical state quantity integral data is cleared by using cross-zero product logic.

[0044] Preferably, the calculation of the comprehensive limiting tension factor based on speed error and acceleration deviation specifically includes:

[0045] The speed error factor is calculated using an exponential function form based on the speed error.

[0046] The acceleration deviation factor is calculated using an exponential function form based on the acceleration deviation.

[0047] The speed error factor and acceleration deviation factor are added together to obtain the comprehensive amplitude limiting tightness factor.

[0048] Preferably, the original torque command is limited and calculated using a comprehensive limiting factor to obtain candidate torque command values, specifically including:

[0049] The difference between the original torque command and the feedback torque command from the previous control cycle is calculated to obtain the original torque increment.

[0050] The scaled boundary is generated by scaling calculation based on the comprehensive limit tension factor and the preset basic allowable increment, and the original torque increment is truncated based on the scaled boundary to obtain the limited torque increment.

[0051] The limited torque increment is added to the feedback torque command from the previous control cycle to obtain the candidate torque command value;

[0052] Release the torque control range when the speed error approaches zero and the acceleration deviation is non-zero, to avoid speed overshoot caused by longitudinal control.

[0053] Preferably, processing the candidate torque command values ​​to obtain the feedback torque command specifically includes:

[0054] The candidate values ​​of the torque command are input into a first-order low-pass filter to obtain the filtered torque command;

[0055] Obtain the hardware torque boundary determined by the physical output limit of the drive motor and the physical braking limit of the air braking system.

[0056] The feedback torque command is obtained by truncating the filtered torque command using hardware torque boundaries.

[0057] A second aspect of the present invention provides a longitudinal control system for new energy heavy-duty trucks based on disturbance observation and feedforward, for implementing the longitudinal control method for new energy heavy-duty trucks based on disturbance observation and feedforward provided in the first aspect of the present invention, comprising:

[0058] The vehicle speed fusion module is used to collect the converted vehicle speed of the driven wheels and the converted vehicle speed of the driving wheels, calculate the slip ratio based on the converted vehicle speed of the driven wheels and the converted vehicle speed of the driving wheels, and process the converted vehicle speed of the driven wheels and the converted vehicle speed of the driving wheels using the slip ratio to obtain the actual vehicle speed and the actual vehicle acceleration.

[0059] The speed planning module is used to receive the target vehicle speed signal, process the target vehicle speed signal, and obtain the reference acceleration and reference vehicle speed.

[0060] The state observation module is used to input the actual vehicle speed, the feedback torque command of the previous control cycle and the brake pedal opening into the third-order linear extended state observer model for calculation to obtain the equivalent compensation torque for slow-varying disturbances.

[0061] The torque synthesis module is used to calculate the feedforward torque based on the reference acceleration, calculate the speed error by using the difference between the reference vehicle speed and the actual vehicle speed, calculate the final feedback torque based on the speed error, and combine the feedforward torque, the final feedback torque and the slow-change disturbance equivalent compensation torque to obtain the original torque command.

[0062] The torque limiting module is used to calculate the acceleration deviation by using the difference between the reference acceleration and the actual vehicle acceleration, calculate the comprehensive limiting factor based on the speed error and acceleration deviation, and use the comprehensive limiting factor to limit the original torque command to obtain the torque command candidate value.

[0063] The post-processing module is used to process the candidate values ​​of the torque command and obtain the feedback torque command.

[0064] The above solution achieves the following beneficial technical effects:

[0065] This invention introduces brake pedal opening, brake pressure build-up time constant, and brake pressure release time constant into a third-order linear extended state observer model. It combines the actual vehicle speed and the feedback torque command from the previous control cycle for discretized iterative calculation, jointly updating the vehicle speed state variable, the slowly varying lumped disturbance state variable, and the transient air braking force equivalent torque state variable. The discretized iterative calculation incorporates the asymmetric fluid pressure build-up time parameter into the model calculation, which can separate road gradient resistance and transient braking force, improving the accuracy of disturbance observation. This provides accurate slowly varying disturbance equivalent compensation torque for the longitudinal control of new energy heavy trucks.

[0066] This invention generates an integral accumulation term by performing conditional integral calculation on the speed error based on multiple preset speed error thresholds when calculating the final feedback torque. It constrains the long-term speed error integral calculation link by using a full-weight accumulation interval, a decay-weight accumulation interval, and a stop accumulation interval divided by the lower limit of the speed error and a preset upper limit. It also uses cross-zero product logic to clear the integral data of historical state quantities. The set conditional integral calculation steps can avoid the integral saturation phenomenon caused by long-term error accumulation and improve the stability of new energy heavy trucks in tracking the target vehicle speed.

[0067] This invention utilizes speed error and acceleration deviation to calculate a comprehensive limiting factor using an exponential function. Based on this factor and a preset allowable increment, a scaling calculation is performed to generate a boundary. The original torque increment is truncated to obtain a limited torque increment. The limiting factor is then used to dynamically adjust the limit boundary of the torque command candidate value. The torque control range is released when the speed error approaches zero and the acceleration deviation is non-zero. This avoids speed overshoot in longitudinal control and ensures the continuity and stability of longitudinal torque output in new energy heavy-duty trucks. Attached Figure Description

[0068] Figure 1 This is a schematic diagram of the system architecture of the present invention;

[0069] Figure 2 This is a schematic diagram of the overall process of the present invention. Detailed Implementation

[0070] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0071] Please see Figure 1 This invention provides a longitudinal control system for new energy heavy trucks based on disturbance observation and feedforward, including: a vehicle speed fusion module, a speed planning module, a state observation module, a torque synthesis module, a torque limiting module, and a post-processing module.

[0072] The longitudinal control system for new energy heavy-duty trucks based on disturbance observation and feedforward is configured in the vehicle controller of the new energy heavy-duty truck. The vehicle controller cyclically executes the calculation process of the vehicle speed fusion module, speed planning module, state observation module, torque synthesis module, torque limiting module, and post-processing module according to a preset control cycle. At the data flow and communication level within the system, the output of the vehicle speed fusion module is connected to the state observation module, torque synthesis module, and torque limiting module to provide the actual vehicle speed and actual vehicle acceleration; the output of the speed planning module is connected to the torque synthesis module and torque limiting module to provide the reference vehicle speed and reference acceleration; the output of the state observation module is connected to the torque synthesis module to provide the observed compensation torque; the output of the torque synthesis module is connected to the torque limiting module to provide the original torque command; the output of the torque limiting module is connected to the post-processing module, which ultimately outputs the feedback torque command to the motor controller for execution.

[0073] See attached document Figure 2 This invention provides a longitudinal control method for new energy heavy-duty trucks based on disturbance observation and feedforward, comprising the following steps:

[0074] S100: The vehicle speed fusion module collects the converted vehicle speeds of the driven wheels and the driven wheels of the new energy heavy truck. The vehicle speed fusion module calculates the slip ratio between the converted vehicle speeds of the driven wheels and the driven wheels. The vehicle speed fusion module adjusts the weighting parameter of the converted vehicle speed of the driven wheels according to the slip ratio. The vehicle speed fusion module performs a weighted calculation on the converted vehicle speeds of the driven wheels and the driven wheels to obtain the original vehicle speed. The vehicle speed fusion module inputs the original vehicle speed into a first-order low-pass filter for processing to obtain the actual vehicle speed. The vehicle speed fusion module performs differential processing on the actual vehicle speed to obtain the actual vehicle acceleration.

[0075] S200, the speed planning module receives the target vehicle speed signal, the speed planning module performs constraint processing on the target vehicle speed according to the preset impact limit value to obtain the reference impact, the speed planning module performs integral calculation on the reference impact and performs constraint processing on the preset acceleration limit value to obtain the reference acceleration, the speed planning module performs integral calculation on the reference acceleration to obtain the reference vehicle speed;

[0076] S300, the operating state observation module establishes a third-order linear extended state observer model. The third-order linear extended state observer model contains three state variables, which are respectively set as vehicle speed state variable, slow-varying lumped disturbance state variable, and transient air braking force equivalent torque state variable. The state observation module obtains the actual vehicle speed, the feedback torque command of the previous control cycle, and the brake pedal opening. The state observation module determines the direction of change of the brake pedal opening. When the brake pedal opening increases, the state observation module selects the brake pressure build-up time constant. When the brake pedal opening decreases, the state observation module selects the brake pressure relief time constant. The state observation module inputs the actual vehicle speed, the feedback torque command of the previous control cycle, and the brake pedal opening into the third-order linear extended state observer model for discretization and iterative calculation to obtain the slow-varying disturbance equivalent compensation torque and the air braking equivalent compensation torque.

[0077] In S400, the torque synthesis module calculates the feedforward torque by multiplying the reference acceleration and the nominal equivalent mass of the new energy heavy truck. The torque synthesis module calculates the speed tracking error by calculating the difference between the reference vehicle speed and the actual vehicle speed. The torque synthesis module calculates the proportional-derivative feedback torque based on the speed tracking error. The torque synthesis module performs conditional integration on the speed tracking error based on multiple preset speed tracking error thresholds to generate an integral accumulation term. When the product of the speed tracking error of the current control cycle and the speed tracking error of the previous control cycle is less than zero, the torque synthesis module clears the integral accumulation term. The torque synthesis module combines the proportional-derivative feedback torque and the integral accumulation term to obtain the final feedback torque. The torque synthesis module adds the feedforward torque, the final feedback torque, and the slow-change disturbance equivalent compensation torque to obtain the original torque command.

[0078] In S500, the torque limiting module calculates the difference between the reference acceleration and the actual vehicle acceleration to obtain the acceleration deviation. The torque limiting module calculates the speed limiting comprehensive limiting looseness factor using the speed tracking error. The torque limiting module calculates the acceleration deviation comprehensive limiting looseness factor using the acceleration deviation. The torque limiting module adds the speed limiting comprehensive limiting looseness factor and the acceleration deviation comprehensive limiting looseness factor to obtain the comprehensive limiting looseness factor. The torque limiting module calculates the difference between the original torque command and the feedback torque command of the previous control cycle to obtain the torque change in the current cycle. The torque limiting module calculates the actual allowable torque change based on the comprehensive limiting looseness factor and the preset basic allowable torque change extreme value. The torque limiting module compares the torque change in the current cycle with the actual allowable torque change for constraint calculation. The torque limiting module adds the constraint calculation result to the feedback torque command of the previous control cycle to obtain the torque command candidate value.

[0079] S600: The post-processing module inputs the candidate value of the torque command into the low-pass filter to obtain the filtered torque command. The post-processing module obtains the physical torque boundary parameters of the motor. The post-processing module uses the physical torque boundary parameters to perform absolute amplitude limit calculation on the filtered torque command to obtain the feedback torque command. The post-processing module sends the feedback torque command to the motor controller for execution.

[0080] See attached document Figure 2 The calculation process of the vehicle speed fusion module may include the following sub-steps:

[0081] S101, the vehicle speed fusion module acquires the converted vehicle speed of the driven wheels and the converted vehicle speed of the driving wheels. For the process of using wheel speed sensors to collect wheel rotation speeds and combining this with the tire's dynamic rolling radius to calculate the converted vehicle speed, those skilled in the art can perform conventional configurations based on vehicle kinematics principles. The specific methods for acquiring the converted vehicle speeds of the driven wheels and the driving wheels are well-known technologies in the field and will not be elaborated here. The vehicle speed fusion module uses the converted vehicle speeds of the driven wheels and the driving wheels to calculate the slip ratio. To prevent division by zero errors due to the driven wheel rotation speed being zero when the vehicle is in a stationary starting state, the vehicle speed fusion module has a preset extremely low vehicle speed threshold, set to 1 km / h. When the converted vehicle speed of the driven wheels is less than or equal to the extremely low vehicle speed threshold, the vehicle speed fusion module directly assigns the slip ratio to zero; when the converted vehicle speed of the driven wheels is greater than the extremely low vehicle speed threshold, the formula for calculating the slip ratio by the vehicle speed fusion module is:

[0082] ;

[0083] in, Indicates the slip ratio; Indicates the vehicle speed converted from the drive wheels; This indicates the speed of the vehicle when the driven wheel is converted.

[0084] S102, the vehicle speed fusion module adaptively adjusts the weighting parameters of the drive wheel speed calculation based on the slip ratio. When a new energy heavy truck starts on a low-traction surface or accelerates with high torque, the drive wheels are prone to macroscopic slippage due to the road surface adhesion. In this case, the rotational speed values ​​collected by the drive wheel sensors will be significantly higher than the vehicle's actual translational speed. To filter out spurious fluctuations in vehicle speed measurement caused by wheel slippage, the vehicle speed fusion module has preset weights for the driven wheels and the drive wheels, with the sum of these weights equal to 1. The module also has a preset slip ratio threshold of 0.15, which corresponds to the typical physical boundary where the adhesion coefficient between the commercial vehicle tire rubber and the road surface reaches its peak. The vehicle speed fusion module compares the calculated slip ratio with the slip ratio threshold. When the slip ratio is greater than the threshold, the module performs dynamic attenuation calculations on the weights of the drive wheels to obtain the actual drive wheel weights. The formula for calculating the actual drive wheel weights is as follows:

[0085] ;

[0086] in, Indicates the actual drive wheel weight; The value in the formula represents the weight of the base drive wheel. The preset attenuation gain coefficient represents the weighting parameter that will be calculated to attenuate to the limit state when the slip ratio exceeds the slip ratio threshold by 0.1 absolute values ​​(i.e., the slip ratio reaches 0.25).

[0087] According to the above calculation formula, when the slip ratio exceeds the limit, the actual drive wheel weight will adaptively decrease as the slip ratio increases. The actual drive wheel weight is set with a lower limit of 0.05. The lower limit of 0.05 is used to prevent data loss caused by the vehicle speed fusion module completely cutting off the vehicle speed conversion signal from the drive wheel. When the slip ratio value is less than or equal to the slip ratio threshold, the vehicle speed fusion module will directly assign the basic drive wheel weight as the actual drive wheel weight.

[0088] S103, the vehicle speed fusion module uses adaptively adjusted weight parameters to perform a weighted fusion calculation of the vehicle speed converted from the driven wheels and the vehicle speed converted from the driving wheels. To ensure the physical correctness of the weighted fusion, after obtaining the actual driving wheel weight, the vehicle speed fusion module simultaneously calculates the actual driven wheel weight. The value of the actual driven wheel weight is equal to a constant 1 minus the value of the actual driving wheel weight. Subsequently, the vehicle speed fusion module uses the actual driven wheel weight and the actual driving wheel weight to perform a weighted calculation. The formula for the vehicle speed fusion module to calculate the original vehicle speed is:

[0089] ;

[0090] in, Indicates the original vehicle speed; This indicates the actual weight of the driven wheel.

[0091] S104, the vehicle speed fusion module inputs the original vehicle speed into an internally set first-order low-pass filter for filtering calculation. The vehicle speed fusion module eliminates high-frequency noise components in the original vehicle speed signal through the first-order low-pass filter. The calculation result output by the first-order low-pass filter is defined as the actual vehicle speed. The actual vehicle speed serves as the real feedback speed used for closed-loop control of the entire new energy heavy truck longitudinal control system. The vehicle speed fusion module further performs differential calculation on the actual vehicle speed in the time dimension. In the specific implementation of the discrete digital control system, the vehicle speed fusion module calculates the difference between the actual vehicle speed of the current control cycle and the actual vehicle speed of the previous control cycle, and divides the difference by the control cycle time to obtain the actual vehicle acceleration of the new energy heavy truck.

[0092] See attached document Figure 2 The calculation process of the velocity planning module may include the following sub-steps:

[0093] S201, the speed planning module receives the target vehicle speed signal. As for the acquisition method of the target vehicle speed signal, those skilled in the art can obtain the relevant signal from the adaptive cruise control system or the intelligent driving domain controller according to the actual configuration of the vehicle. The specific acquisition method is a well-known technology in the field and will not be described in detail here.

[0094] S202, the speed planning module obtains the original impact value based on the changing trend of the target vehicle speed signal. Specifically, the speed planning module uses a discrete difference algorithm to perform two consecutive difference calculations on the target vehicle speed signal in the time dimension, and divides the calculation result by the square of the control cycle time to obtain a physical quantity characterizing the rate of change of the vehicle's longitudinal acceleration, i.e., the original impact value. The speed planning module internally sets an impact value limit, specifically set at ±2.0 m / s3. This value is determined based on the physical torsional clearance of the new energy heavy truck transmission system and the passenger comfort boundary. The speed planning module uses the impact value limit to perform absolute amplitude truncation processing on the original impact value to obtain a reference impact value. The underlying principle of limiting the impact value is that the impact value represents the rate of change of acceleration in kinematics, and it directly corresponds to the rate of change of torque of the drive motor in dynamics. By limiting this value, it is possible to prevent the new energy heavy truck from generating large longitudinal mechanical jerks at the moment of switching between acceleration and deceleration.

[0095] S203, the speed planning module performs integral calculations on the reference impact over time to obtain a temporary reference acceleration. The speed planning module has an internal acceleration limit value, specifically set to ±0.8 m / s². This value is determined based on the peak output power of the drive motor of the new energy heavy truck under a full load of 49 tons and the physical boundary of the maximum tire adhesion. The speed planning module uses the acceleration limit value to truncate the temporary reference acceleration to obtain the reference acceleration. The principle of limiting the acceleration value is to avoid the new energy heavy truck generating power demands that exceed physical limits due to a step change in the target vehicle speed, thereby preventing the drive motor from overheating or the drive wheels from experiencing macroscopic slippage.

[0096] S204, the speed planning module performs integral calculation of the reference acceleration in the time dimension to obtain the temporary reference vehicle speed. The speed planning module performs speed limiting processing on the temporary reference vehicle speed according to the maximum driving speed set by the vehicle to obtain the reference vehicle speed of the current control cycle. Through the above cascaded limiting calculation process, the speed planning module transforms the target vehicle speed signal that is prone to step jumps into a reference vehicle speed signal with a smooth transition, and synchronously outputs the reference acceleration and reference impact that are perfectly matched with the reference vehicle speed in terms of physical kinematics. The reference acceleration is used as the basic input variable for subsequent feedforward torque calculation.

[0097] See attached document Figure 2 The calculation process of the state observation module may include the following sub-steps:

[0098] S301, the state observation module establishes a third-order linear extended state observer model. Traditional linear extended state observers aggregate unmodeled internal dynamics and external disturbances into a single scalar for estimation. Due to the bandwidth limitation of a single observer, this method cannot simultaneously and accurately track the slowly changing road slope resistance and the rapidly responding air braking force. Considering the physical characteristics of the air braking system of new energy heavy trucks, where the response speed is significantly different from slowly changing disturbances such as road slope, the state observation module independently separates the air braking force from the total disturbance. The third-order linear extended state observer model contains three state variables, which are respectively set as vehicle speed state variable, slowly varying lumped disturbance state variable, and transient air braking force equivalent torque state variable. The slowly varying lumped disturbance state variable covers the equivalent low-frequency disturbance caused by changes in road slope resistance, rolling resistance, and vehicle load mass.

[0099] S302, the state observation module acquires the brake pedal opening signal and determines the direction of change in the brake pedal opening. The third-order linear extended state observer includes an asymmetric first-order air brake dynamics model. The air brake system of new energy heavy trucks exhibits asymmetric characteristics in fluid dynamics during the inflation and depressurization processes. The inflation and depressurization process requires overcoming the pipeline volume and valve body mechanical resistance, which takes a relatively long time. The depressurization process, on the other hand, requires a shorter time because the gas is directly discharged to the atmosphere. When the brake pedal opening increases, the state observation module determines that it is in the pressure building process and selects a brake pressure building time constant, specifically set to 0.35 seconds. When the brake pedal opening decreases, the state observation module determines that it is in the depressurization process and selects a brake depressurization time constant, specifically set to 0.03 seconds. The brake pressure building time constant is greater than the brake depressurization time constant.

[0100] S303, the state observation module acquires the actual vehicle speed of the current control cycle and the feedback torque command of the previous control cycle. The state observation module inputs the actual vehicle speed, the feedback torque command of the previous control cycle, and the brake pedal opening into the third-order linear extended state observer model for discretization and iterative calculation. The formula for the state observation module to perform iterative calculation based on the forward Euler method is as follows:

[0101] ;

[0102] ;

[0103] ;

[0104] in, This represents the estimated value of the vehicle speed state variable for the next control cycle; This represents the estimated value of the vehicle speed state variable for the current control cycle; This represents the estimated value of the slowly varying lumped disturbance state variable for the next control cycle; This represents the estimated value of the slowly varying lumped disturbance state variable for the current control cycle; This represents the estimated state variable of the transient gas braking force equivalent torque for the next control cycle; This represents the estimated state variable of the transient air braking force equivalent torque during the current control cycle. This indicates the control cycle time parameter; This indicates the feedback torque command from the previous control cycle; This indicates the currently selected air braking time constant; This represents the braking force mapping gain, which is used to map the brake pedal opening value to the equivalent motor torque. The specific value of this gain can be directly obtained by those skilled in the art through bench calibration tests of the vehicle air brake system. This indicates the brake pedal opening, with a value ranging from 0 to 1; This represents the speed estimation error, calculated as follows: , Indicates the actual vehicle speed; The system control gain is represented by the following calculation method: ; This indicates the nominal equivalent mass of the new energy heavy truck. The nominal equivalent mass of the new energy heavy truck under full load is specifically set at 26,250 kg. It is calculated by multiplying the full load design mass of the vehicle (25,000 kg) by the equivalent inertia amplification factor of the rotating components of the transmission system (1.05). This indicates the overall speed ratio of the transmission system in a new energy heavy-duty truck. Indicates the dynamic rolling radius of the wheels of a new energy heavy-duty truck; , and This represents the observer gain parameter, and the observer gain parameter is calculated as follows: , , , This indicates the observer bandwidth. The observer bandwidth is set to 10 rad / s, which is an engineering application parameter that takes into account the observer's response speed to external disturbances and its ability to suppress and attenuate high-frequency sensor measurement noise.

[0105] S304. After the iterative calculation is completed, the state observation module extracts the corresponding state variables to obtain the compensation torque result. The state observation module multiplies the estimated value of the slow-varying lumped disturbance state variable of the current control cycle by the nominal equivalent mass to obtain the slow-varying disturbance equivalent compensation torque. The state observation module directly outputs the estimated value of the transient air braking force equivalent torque state variable of the current control cycle as the air braking equivalent compensation torque. The state observation module uses a single observer to simultaneously realize the estimation processing of vehicle speed, lumped disturbance and nonlinear braking force.

[0106] See attached document Figure 2 The calculation process of the torque synthesis module may include the following sub-steps:

[0107] S401, the torque synthesis module acquires the reference acceleration output by the speed planning module and the slow-variable disturbance equivalent compensation torque output by the state observation module. The torque synthesis module acquires the nominal equivalent mass, wheel dynamic rolling radius, and overall transmission ratio of the new energy heavy truck. Based on the reference acceleration generated by the planning, the torque synthesis module calculates the feedforward torque. The torque synthesis module multiplies the reference acceleration by the nominal equivalent mass to obtain the feedforward traction force, and then multiplies the feedforward traction force by the wheel dynamic rolling radius and divides it by the overall transmission ratio to obtain the feedforward torque. The formula for calculating the feedforward torque by the torque synthesis module is:

[0108] ;

[0109] in, Indicates feedforward torque; Indicates the nominal equivalent mass; Indicates the reference acceleration for the current control cycle; This indicates the overall speed ratio of the transmission system in a new energy heavy-duty truck. This indicates the dynamic rolling radius of the wheels of a new energy heavy-duty truck.

[0110] Feedforward torque represents the basic physical quantity of power required to drive a new energy heavy truck to achieve the planned acceleration. Through this direct calculation of the product of physical quantities, the torque synthesis module enables the generated torque command to respond in advance relative to the actual vehicle speed deviation, thereby reducing the system's dependence on subsequent feedback control adjustments.

[0111] S402, the torque synthesis module obtains the reference vehicle speed and the actual vehicle speed for the current control cycle. The torque synthesis module subtracts the actual vehicle speed from the reference vehicle speed to obtain the speed error for the current control cycle, i.e., calculates... ,in For reference speed; This represents the actual vehicle speed. The torque synthesis module utilizes the speed error to perform proportional-derivative feedback control calculations.

[0112] For the specific numerical matching of proportional gain and derivative gain, those skilled in the art can obtain the tuning results by using the conventional critical proportional gain method combined with actual vehicle bench calibration tests, based on the longitudinal dynamic transmission characteristics of the vehicle. The formula for calculating the proportional-derivative feedback torque of the torque synthesis module is as follows:

[0113] ;

[0114] in, This indicates the proportional-derivative feedback torque; Indicates proportional gain; Represents differential gain; This indicates the speed error of the current control cycle; This indicates the speed error of the previous control cycle; This indicates the control cycle time parameter.

[0115] S403, the torque synthesis module performs conditional integral calculations within a wide speed error range. Conventional unrestricted integral processes are prone to long-term integral saturation accumulation when encountering the large inertia characteristics of new energy heavy trucks, which can lead to speed tracking overshoot. To solve this problem, the torque synthesis module sets three-level integral thresholds. The torque synthesis module internally sets a lower limit for speed error and a preset upper limit. The lower limit for speed error is specifically set to 1 km / h, which corresponds to the unavoidable steady-state small disturbance zone when the vehicle is driving on a normal road. The preset upper limit is specifically set to 4 km / h, which corresponds to the critical boundary to prevent integral saturation effect caused by a step increase in the target vehicle speed.

[0116] The torque synthesis module determines the absolute value of the speed error in the current control cycle. When the absolute value of the speed error is less than the lower limit of the speed error, the torque synthesis module adds the speed error to the internal integral accumulation term with a full weight of 1.

[0117] When the absolute value of the speed error is greater than or equal to the lower limit of the speed error and less than the preset upper limit, the torque synthesis module determines that the speed error is within a wide speed error range. Within this range, the integral weight decreases linearly with the increase of the absolute value of the speed error. The torque synthesis module adds the speed error multiplied by the decrease weight to the integral accumulation term.

[0118] When the absolute value of the speed error is greater than or equal to the preset upper limit, the torque synthesis module determines that the error exceeds the preset upper limit. At this time, the torque synthesis module freezes the integral accumulation term directly instead of clearing it to zero. The freezing operation means that the torque synthesis module keeps the integral accumulation term at its current value and no longer accumulates the speed error of the current control cycle. This is intended to both retain the effective compensation amount accumulated in the early stage and cut off the erroneous accumulation path caused by subsequent large deviations. The formula for the torque synthesis module to calculate the update of the integral accumulation term is:

[0119] ;

[0120] in, This represents the integral accumulation term after the current control cycle update; This represents the integral accumulation term from the previous control cycle; This represents the integral weight. Based on the aforementioned logic, when... hour, ;when hour, ;when hour, .

[0121] Meanwhile, the torque synthesis module uses zero-crossing error as the condition for integral zeroing. The only condition for integral zeroing is that the product of the speed error of the current control cycle and the speed error of the previous control cycle is negative (i.e., ...). When this unique condition is met, it means that the actual vehicle speed has just crossed the reference vehicle speed. The torque synthesis module immediately forces the integral accumulation term to be zero. The error zero-crossing and zeroing mechanism ensures that the direction of the force of the integral value is always consistent with the current control requirements for reducing the error, and avoids the vehicle speed overshoot caused by the accumulation of historical error integral values.

[0122] S404, the torque synthesis module adds the proportional-derivative feedback torque to the integral result to obtain the final feedback torque. The calculation formula is as follows:

[0123] ;

[0124] in, This indicates the final feedback torque; This indicates the preset integral gain.

[0125] Subsequently, the torque synthesis module sums the feedforward torque, the final feedback torque, and the equivalent compensation torque for the slowly varying disturbance to obtain the original torque command. The formula for calculating the original torque command by the torque synthesis module is as follows:

[0126] ;

[0127] in, Indicates the original torque command; Indicates feedforward torque; This indicates the proportional-derivative feedback torque; This represents the integral accumulation term after the current control cycle update; This represents the equivalent compensation torque for slow-varying disturbances.

[0128] See attached document Figure 2 The calculation process of the torque limiting module may include the following sub-steps:

[0129] S501, the torque limiting module obtains the original torque command output by the torque synthesis module, obtains the reference acceleration output by the speed planning module, and obtains the actual vehicle acceleration. The torque limiting module calculates the difference between the reference acceleration and the actual vehicle acceleration to obtain the acceleration deviation. The formula for calculating the acceleration deviation by the torque limiting module is:

[0130] ;

[0131] in, This indicates the acceleration deviation during the current control cycle; Indicates the reference acceleration for the current control cycle; It represents the actual vehicle acceleration in the current control cycle; acceleration deviation is a physical advance reflection of the vehicle speed change trend. Using acceleration deviation as a forward-looking indicator helps the controller identify potential vehicle speed overshoot risks in advance and respond accordingly.

[0132] S502, the torque limiting module acquires the speed error transmitted by the torque synthesis module. The torque limiting module calculates a speed error factor based on the speed error and an acceleration deviation factor based on the acceleration deviation. To achieve nonlinear control gain adjustment, both the speed error factor and the acceleration deviation factor are designed using exponential functions. Exponential functions provide a smooth numerical transition when the deviation is small and quickly reach numerical saturation when the deviation increases. The formulas for calculating the speed error factor and the acceleration deviation factor by the torque limiting module are as follows:

[0133] ;

[0134] ;

[0135] in, Indicates the speed error factor; Indicates the acceleration deviation factor; The constant in the formula represents the speed error of the current control cycle. and This determines that the speed error factor will vary from 0.4 to 1.0; This indicates the acceleration deviation during the current control cycle;

[0136] To avoid ambiguity in the reference of constant parameters, their mathematical function is explained in detail here: the constants 1.0 and 0.6 in the velocity error factor formula are amplitude offset coefficients, which together determine the velocity error factor. The variation range is from 0.4 to 1.0; the constant 0.5 at the beginning of the acceleration deviation factor formula is the amplitude gain coefficient, which determines the acceleration deviation factor. The range of variation is from 0 to 0.5; while The constant 0.5 in the denominator, and The constant 0.15 in the figure refers to the shape coefficients of the convergence rate of the control function curve. The specific values ​​of the shape coefficients are engineering application parameters obtained through the calibration of the dynamic response characteristics of new energy heavy trucks.

[0137] S503, the torque limiting module superimposes the speed error factor and the acceleration deviation factor to calculate the comprehensive limiting tightness factor for torque increment limiting. The formula for calculating the comprehensive limiting tightness factor by the torque limiting module is:

[0138] ;

[0139] in, This represents the overall limiting factor, a constant. Used to truncate the upper limit of the comprehensive limit factor to prevent the value from exceeding the limit.

[0140] Based on the above calculation logic, the comprehensive limiting tightness factor of torque increment limiting is obtained by superimposing the speed error factor and the acceleration deviation factor. The acceleration deviation factor increases monotonically with the increase of the absolute value of acceleration deviation in the form of an exponential function. Common torque limiting strategies only tighten the limiting based on speed error. When the speed error approaches zero, the limitation will be too strict. However, in the embodiments of this application, due to the introduction of the acceleration deviation factor, when the speed error approaches zero but the acceleration deviation is obvious, the limiting remains in a relaxed state to allow the torque to adjust quickly, thereby suppressing overshoot.

[0141] S504, the torque limiting module dynamically limits the change in the original torque command using a comprehensive limiting tightness factor. The torque limiting module has a preset basic allowable increment, the positive and negative boundary values ​​of which are determined based on the maximum torque response rate of the drive motor and the smoothness requirements of the vehicle's transmission system. The torque limiting module calculates the difference between the original torque command and the feedback torque command from the previous control cycle to obtain the original torque increment. The torque limiting module then scales the basic allowable increment using the comprehensive limiting tightness factor and truncates the original torque increment based on the scaled boundary. Finally, the truncated result is added to the feedback torque command from the previous control cycle to obtain a candidate torque command value. The formula for calculating the candidate torque command value by the torque limiting module is:

[0142] ;

[0143] ;

[0144] ;

[0145] in, Indicates the original torque increment; This indicates the initial torque command for the current control cycle; This indicates the feedback torque command from the previous control cycle; This indicates the torque increment after the limit is applied; This represents the positive boundary value of the allowable increment. This indicates the candidate value for the torque command in the current control cycle.

[0146] After obtaining the candidate torque command value for the current control cycle, the S505 torque limiting module transmits it as the final torque limiting result to the post-processing module for further filtering, noise reduction, and hardware boundary protection calculations.

[0147] See attached document Figure 2 The post-processing module's calculation process may include the following sub-steps:

[0148] S601, the post-processing module obtains the candidate value of the torque command output by the torque limiting module. Torque command candidate values After the aforementioned two-factor dynamic limiting calculation, although a large sudden change that would cause vehicle speed overshoot is avoided, the candidate value signal still contains high-frequency jitter components due to the unavoidable high-frequency measurement noise of the sensors in the front-end actual vehicle speed and actual acceleration measurement links. Directly sending the torque command with high-frequency jitter to the underlying actuator will cause the motor output shaft to oscillate at high frequency, thereby affecting the mechanical life of the transmission system.

[0149] S602, the post-processing module processes torque command candidate values. A first-order low-pass filter is performed. The post-processing module is equipped with a first-order low-pass filter with a cutoff frequency of 10 Hz. The physical principle behind setting this cutoff frequency is that the effective longitudinal dynamic response bandwidth of a large inertial system like a new energy heavy truck is relatively narrow, usually much lower than 10 Hz. Meanwhile, the white noise introduced by the chassis speed sensor has a high frequency. Using 10 Hz as the cutoff frequency can filter out high-frequency noise components and ensure that the effective control command signal will not generate excessive phase delay that could cause instability in the closed-loop system during transmission. The specific digital difference equation for the first-order low-pass filter can be configured by those skilled in the art based on conventional signal processing theory. This is a well-known technology in the field and will not be elaborated here.

[0150] S603, after filtering, the post-processing module performs amplitude limiting on the filtered torque signal. The post-processing module obtains the hardware torque boundary under the current vehicle operating state. Specifically, the hardware torque boundary is determined by the physical output limit of the drive motor and the physical braking limit of the air brake system. The physical output limits of the drive motor for forward drive and reverse feedback are determined by the motor's external characteristic curve at real-time speed. The physical braking limit of the air brake system is determined by the maximum braking torque parameter of the basic air brake. The post-processing module uses the aforementioned hardware torque boundary to truncate the filtered torque signal at upper and lower limits to obtain the feedback torque command. The amplitude limiting operation prevents the output of over-limit instructions due to any abnormal calculations in the upper-level algorithm.

[0151] S604, the after-processing module will feed back torque commands. The power is transmitted to the motor controller of the new energy heavy truck for power execution, and a torque command is received upon feedback. When there is a negative braking demand that exceeds the physical limit of the motor's maximum energy recovery, the post-processing module simultaneously sends the corresponding mechanical braking command to the vehicle's air brake controller for mechanical braking compensation. Through the cascaded processing of first-order low-pass filtering and amplitude limiting, the post-processing module outputs a smooth final torque control quantity that conforms to the underlying hardware safety physical boundary.

[0152] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A longitudinal control method for new energy heavy-duty trucks based on disturbance observation and feedforward, characterized in that, include: The system collects the converted vehicle speeds of the driven wheels and the driving wheels, calculates the slip ratio based on the converted vehicle speeds of the driven wheels and the driving wheels, and processes the converted vehicle speeds of the driven wheels and the driving wheels using the slip ratio to obtain the actual vehicle speed and the actual vehicle acceleration. Receive the target vehicle speed signal, process the target vehicle speed signal, and obtain the reference acceleration and reference vehicle speed; The actual vehicle speed, the feedback torque command from the previous control cycle, and the brake pedal opening are input into the third-order linear extended state observer model for calculation to obtain the equivalent compensation torque for slow-varying disturbances. The feedforward torque is calculated based on the reference acceleration, the speed error is obtained by calculating the difference between the reference vehicle speed and the actual vehicle speed, the final feedback torque is calculated based on the speed error, and the original torque command is obtained by merging the feedforward torque, the final feedback torque and the slow-change disturbance equivalent compensation torque. The acceleration deviation is obtained by calculating the difference between the reference acceleration and the actual vehicle acceleration. A comprehensive limiting factor is calculated based on the speed error and the acceleration deviation. The original torque command is then limited by the comprehensive limiting factor to obtain a candidate value for the torque command. The candidate values ​​of the torque command are processed to obtain the feedback torque command.

2. The longitudinal control method for new energy heavy trucks based on disturbance observation and feedforward according to claim 1, characterized in that, The process of calculating the vehicle speed from the driven wheel and the driving wheel using the slip ratio to obtain the actual vehicle speed and actual vehicle acceleration specifically includes: The slip ratio is compared with a preset slip ratio threshold; When the slip ratio is greater than the slip ratio threshold, an attenuation calculation is performed on the preset basic drive wheel weight to obtain the actual drive wheel weight; When the slip ratio is less than or equal to the slip ratio threshold, the weight of the basic drive wheel is directly assigned to the weight of the actual drive wheel; Obtain a preset constant 1, calculate the difference between the constant 1 and the actual driving wheel weight, and obtain the actual driven wheel weight; The original vehicle speed is obtained by weighting the converted vehicle speed of the driven wheel and the converted vehicle speed of the driven wheel using the actual driven wheel weight and the actual driving wheel weight; The original vehicle speed is low-pass filtered to obtain the actual vehicle speed, and the actual vehicle speed is differentiated to obtain the actual vehicle acceleration.

3. The longitudinal control method for new energy heavy-duty trucks based on disturbance observation and feedforward according to claim 1, characterized in that, Processing the target vehicle speed signal to obtain the reference acceleration and reference vehicle speed specifically includes: The original impact intensity is obtained based on the changing trend of the target vehicle speed signal; The original impact is constrained according to a preset impact limit value to obtain a reference impact. The temporary reference acceleration is obtained by integrating the reference impact, and the temporary reference acceleration is constrained according to a preset acceleration limit value to obtain the reference acceleration. The reference acceleration is integrated to obtain a temporary reference speed, and the temporary reference speed is speed-limited according to the set maximum driving speed to obtain the reference speed.

4. The longitudinal control method for new energy heavy trucks based on disturbance observation and feedforward according to claim 1, characterized in that, The actual vehicle speed, the feedback torque command from the previous control cycle, and the brake pedal opening are input into a third-order linear extended state observer model for calculation to obtain the equivalent compensation torque for slow-varying disturbances. Specifically, this includes: The third-order linear extended state observer model includes vehicle speed state variables, slow-varying lumped disturbance state variables, and transient air braking force equivalent torque state variables. Determine the direction of change in the brake pedal opening; When the brake pedal opening increases, a preset brake pressure build-up time constant is selected; When the brake pedal opening decreases, a preset brake pressure relief time constant is selected; The actual vehicle speed, the feedback torque command of the previous control cycle, and the brake pedal opening are used as input items. Combined with the selected brake pressure build-up time constant or brake pressure relief time constant, discretization iterative calculation is performed in the third-order linear extended state observer model based on the forward Euler method. The vehicle speed state variable, the slow-varying lumped disturbance state variable, and the transient air braking force equivalent torque state variable are updated simultaneously. Extract the updated estimated value of the slow-varying lumped disturbance state variable, multiply the extracted estimated value of the slow-varying lumped disturbance state variable by the nominal equivalent mass, and obtain the slow-varying disturbance equivalent compensation torque.

5. The longitudinal control method for new energy heavy-duty trucks based on disturbance observation and feedforward according to claim 1, characterized in that, The calculation of the final feedback torque based on the speed error specifically includes: Calculate the proportional-derivative feedback torque based on the speed error; The speed error is conditionally integrated based on multiple preset speed error thresholds to generate an integral accumulation term; The final feedback torque is obtained by combining the proportional-derivative feedback torque and the integral accumulation term.

6. The longitudinal control method for new energy heavy-duty trucks based on disturbance observation and feedforward according to claim 5, characterized in that, The step of performing conditional integration on the speed error based on multiple preset speed error thresholds to generate an integral accumulation term specifically includes: Determine the absolute value of the speed error in the current control cycle; When the absolute value of the speed error is less than the lower limit of the speed error, the speed error is added to the integral accumulation term with full weight. When the absolute value of the speed error is greater than or equal to the lower limit of the speed error and less than the preset upper limit, the speed error multiplied by the attenuation weight is added to the integral accumulation term; When the absolute value of the speed error is greater than or equal to the preset upper limit, the integral accumulation term is kept at the current value, and the accumulation of the speed error for the current control cycle is stopped; When the product of the speed error in the current control cycle and the speed error in the previous control cycle is less than zero, the integral accumulation term is cleared.

7. The longitudinal control method for new energy heavy trucks based on disturbance observation and feedforward according to claim 1, characterized in that, The calculation of the comprehensive amplitude limiting tightness factor based on the speed error and the acceleration deviation specifically includes: The speed error factor is calculated using an exponential function based on the aforementioned speed error. The acceleration deviation factor is calculated using an exponential function form based on the aforementioned acceleration deviation. The combined amplitude limiting factor is obtained by adding the speed error factor and the acceleration deviation factor.

8. The longitudinal control method for new energy heavy trucks based on disturbance observation and feedforward according to claim 1, characterized in that, The process of using the comprehensive limiting factor to limit the original torque command and obtaining candidate torque command values ​​specifically includes: The difference between the original torque command and the feedback torque command of the previous control cycle is calculated to obtain the original torque increment; The scaled boundary is generated by scaling calculation based on the comprehensive limiting tension factor and the preset basic allowable increment, and the original torque increment is truncated based on the scaled boundary to obtain the limited torque increment. The torque increment after the limitation is added to the feedback torque command of the previous control cycle to obtain the candidate value of the torque command.

9. The longitudinal control method for new energy heavy trucks based on disturbance observation and feedforward according to claim 1, characterized in that, Processing the candidate torque command values ​​to obtain the feedback torque command specifically includes: The candidate values ​​of the torque command are input into a first-order low-pass filter to obtain the filtered torque command; Obtain the hardware torque boundary determined by the physical output limit of the drive motor and the physical braking limit of the air braking system. The feedback torque command is obtained by truncating the filtered torque command using the hardware torque boundary.

10. A longitudinal control system for new energy heavy-duty trucks based on disturbance observation and feedforward, characterized in that, The method for implementing the longitudinal control of new energy heavy-duty trucks based on disturbance observation and feedforward as described in any one of claims 1-9 includes: The vehicle speed fusion module is used to collect the converted vehicle speed of the driven wheels and the converted vehicle speed of the driving wheels, calculate the slip ratio based on the converted vehicle speed of the driven wheels and the converted vehicle speed of the driving wheels, and process the converted vehicle speed of the driven wheels and the converted vehicle speed of the driving wheels using the slip ratio to obtain the actual vehicle speed and the actual vehicle acceleration. The speed planning module is used to receive the target vehicle speed signal, process the target vehicle speed signal, and obtain the reference acceleration and reference vehicle speed. The state observation module is used to input the actual vehicle speed, the feedback torque command of the previous control cycle and the brake pedal opening into the third-order linear extended state observer model for calculation to obtain the equivalent compensation torque for slow-varying disturbances. The torque synthesis module is used to calculate the feedforward torque based on the reference acceleration, calculate the speed error by using the difference between the reference vehicle speed and the actual vehicle speed, calculate the final feedback torque based on the speed error, and combine the feedforward torque, the final feedback torque and the slow-change disturbance equivalent compensation torque to obtain the original torque command. The torque limiting module is used to calculate the acceleration deviation by using the difference between the reference acceleration and the actual vehicle acceleration, calculate a comprehensive limiting factor based on the speed error and the acceleration deviation, and use the comprehensive limiting factor to limit the original torque command to obtain a torque command candidate value. The post-processing module is used to process the candidate values ​​of the torque command and obtain the feedback torque command.

Images