Redundant brake control method and storage medium for an autonomous vehicle

Abstract

The application discloses a kind of redundancy brake control method and storage medium of automatic driving vehicle, and is applied to vehicle controller.The method is detected when automatic driving module, collision warning or remote monitoring triggers brake demand, and the target deceleration is calculated in combination with vehicle speed and obstacle distance.By collecting actual acceleration, parking air chamber air pressure and motor torque, total braking force is estimated and electronic parking brake force is separated, and then road adhesion coefficient is calculated.In redundancy brake mode, according to target deceleration, adhesion condition and the upper limit of the ability of two systems, a performance index containing deceleration deviation and parking penalty term is constructed, and the target motor deceleration and the target electronic parking deceleration are obtained by joint distribution.Finally, using the deceleration air pressure mapping table containing online correction amount, the target parking deceleration is converted into target air pressure, and the air pressure is adjusted through closed-loop control to realize precise and stable redundancy driving brake.

Description

Technical Field

[0001] This application relates to the field of vehicle braking control for autonomous vehicles, and in particular to a redundant braking control method based on the coordination of electric motor braking and electronic parking system. Background Technology

[0002] When autonomous vehicles operate in real-world traffic environments, they may face complex conditions such as insufficient response of the braking system, reduced braking capacity of the electric motor due to temperature or battery power limitations, abnormal sensor signals, and sudden changes in road surface adhesion coefficient. Therefore, to ensure that vehicles can maintain safe braking capabilities continuously when performing autonomous driving tasks, they need to be equipped with multi-level redundant safety braking strategies. This allows them to utilize other braking actuators, such as the electronic parking brake (EPB), to achieve stable and reliable deceleration or stopping even if the main braking system fails or becomes insufficient.

[0003] In existing technologies, electronic parking brake (EPB) systems for commercial vehicles (with pneumatic brakes) are typically used for vehicle parking and lack precise braking control capabilities tailored to different driving conditions. Furthermore, the EPB braking response is strongly correlated with factors such as road surface adhesion, load variations, and pneumatic control errors. When used in conjunction with electric motor braking as a redundant braking method, the lack of accurate braking force estimation and adaptive control strategies to suit different adhesion conditions may lead to risks of under-braking or over-braking. Additionally, there is a lack of robust safety degradation control methods in case of sensor failure or unreliable braking capacity estimation.

[0004] Therefore, how to construct a redundant braking method that can maintain stable response under complex working conditions, achieve reasonable joint distribution of motor braking and EPB, realize adaptive braking control according to different road surfaces, and have the ability to degrade safety in the event of sensor failure has become a problem that needs to be solved in the field of braking safety of autonomous vehicles. Summary of the Invention

[0005] This invention aims to address the problem of unreliable braking in autonomous vehicles operating in complex road environments, caused by factors such as limited motor braking capacity, significant changes in road surface adhesion conditions, and sensor malfunctions. To improve the redundant braking capability of autonomous vehicles in emergency situations and braking system failure scenarios, this invention proposes a redundant braking control method based on the coordinated control of motor braking and the electronic parking brake system. Through comprehensive modeling of the target deceleration, road surface adhesion conditions, upper limit of braking capacity, and air pressure mapping relationship of the electronic parking brake system, the method optimizes the allocation of redundant braking strategies, adaptively adjusts them, and handles fault degradation, ensuring that the vehicle can stably achieve the desired braking deceleration under various operating conditions.

[0006] The redundant braking control method of this invention is applied to the vehicle controller of an autonomous vehicle, including an electric motor braking system and an electronic parking brake (EPB) system. First, when the autonomous driving module, collision warning system, or remote monitoring system triggers a braking demand, the vehicle controller acquires the current vehicle speed, target vehicle speed, and distance to the obstacle ahead, and calculates the required target deceleration of the vehicle based on a preset safe distance. When the distance is greater than the safe distance, the target deceleration is calculated based on the vehicle's kinematics and limited; when the obstacle distance is less than the safe distance, the maximum permissible deceleration is directly used as the target deceleration to ensure braking safety.

[0007] In estimating vehicle braking capacity, the controller calculates the current total braking force based on the vehicle's longitudinal acceleration, EPB parking spring chamber pressure, and motor braking torque information, using the vehicle's mass. This is then subtracted from the braking force generated by the motor braking to obtain an estimate of the actual braking force of the EPB. Furthermore, the vehicle controller calculates the ratio of the estimated total braking force to the vehicle's equivalent normal load to obtain the current road surface adhesion coefficient and equivalent adhesion capacity. This adhesion capacity serves as a crucial basis for determining the upper limit of the maximum usable deceleration of the EPB.

[0008] When redundant braking demand is identified, the controller comprehensively considers the target deceleration, road adhesion conditions, and the availability of motor braking and EPB under the current operating conditions. It constructs a performance index function with the total deceleration deviation and EPB usage penalty as the target, and then optimizes the solution to jointly allocate the target deceleration between motor braking and EPB. In the allocation strategy, if the detected target deceleration is greater than a preset emergency braking threshold, priority is given to ensuring that the combined deceleration of the motor and EPB is not lower than the target deceleration; if the target deceleration is lower, the EPB usage penalty weight is increased so that the motor braking undertakes most of the braking task, thereby improving the smoothness and economy of the vehicle's braking.

[0009] For EPB (Electronic Braking Brake) control, this invention employs a "deceleration-pressure mapping table" containing basic calibration data and online correction values ​​to convert the target EPB deceleration into a corresponding target air pressure. The vehicle controller implements closed-loop regulation of the EPB air pressure based on real-time feedback signals from the air pressure sensor, gradually bringing the actual air pressure closer to the target value to ensure that the actual deceleration contributed by the EPB meets the allocation results. Furthermore, the controller adaptively updates the online correction values ​​in the mapping table by comparing the difference between the actual EPB deceleration and the target deceleration in real time, according to a preset learning rate. Amplitude constraints ensure stable convergence of the correction values, enabling the mapping table to automatically adjust to a control relationship that better reflects the actual braking effect under different load conditions, brake fade states, and changes in road surface adhesion.

[0010] Meanwhile, this invention also proposes a mechanism for monitoring and adjusting the braking effect. The vehicle controller continuously collects the vehicle's longitudinal acceleration within a preset time window, calculates the actual total deceleration, and compares it with the target deceleration. When the actual deceleration is consistently lower than a set proportion of the target deceleration, the system automatically increases the target deceleration of the EPB or adjusts the target air pressure to enhance its braking effect; when the actual deceleration is consistently higher than the corresponding proportion threshold, the contribution of the EPB is reduced to avoid excessive deceleration causing ride discomfort or vehicle instability.

[0011] When a sensor malfunctions or the estimated adhesion coefficient exceeds a reliable range, this invention further incorporates an algorithm degradation strategy. After degradation, the system suspends online correction updates and switches to a fixed allocation method based on a preset ratio or priority to avoid outputting unreasonable braking commands under abnormal conditions. Once the fault is resolved, the system resumes its normal optimized allocation strategy and online correction mechanism, thereby ensuring the continuity and safety of redundant braking functions.

[0012] Furthermore, this invention proposes a mechanism for maintaining multiple sets of online correction values ​​based on different road surface adhesion coefficient ranges. By dividing the road surface adhesion range into multiple ranges and assigning different correction value sets, the system can establish differentiated EPB deceleration pressure mapping tables under various typical road surface conditions such as dry, wet, and icy conditions, achieving adaptive capability for multiple road conditions. When the road surface adhesion switches from one range to another, the system automatically selects and updates the corresponding correction value set for that range, ensuring that braking control maintains optimal performance under all road conditions.

[0013] Finally, the present invention also proposes a computer-readable storage medium in which the stored program can implement all the steps of the above-described redundant braking control method when executed by a processor, providing software-level support for the implementation of the present invention in a vehicle control system.

[0014] Through the above series of technical solutions, this invention enables autonomous vehicles to stably and accurately perform redundant braking control even when the main braking system is limited, road conditions are complex, or sensors are abnormal, thereby significantly improving the braking safety, reliability, and environmental adaptability of vehicles in autonomous driving environments. Attached Figure Description

[0015] Figure 1 This is a schematic diagram illustrating the calculation of target deceleration based on obstacle distance and safety distance in an embodiment of this application.

[0016] Figure 2 This is a schematic diagram of the hardware framework of the redundant braking control system of the present invention.

[0017] Figure 3 This is a schematic diagram of the main steps of the redundant braking control method of the present invention. Detailed Implementation

[0018] The redundant braking control method of this application will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that the schematic structures in the drawings are only used to explain the technical principles of this application, and their proportions, directions, and logical relationships are not necessarily drawn according to actual structures. Those skilled in the art can adjust or modify them according to actual needs after reading this specification, but this should not be construed as a limitation on the scope of protection of this application.

[0019] like Figure 1 As shown in the figure, this embodiment uses a typical driving scenario where an autonomous vehicle faces an obstacle in front as an example to illustrate the vehicle's current speed v and target speed. Distance d of the obstacle ahead and preset safety distance The relationship between the current vehicle speed v and the target vehicle speed is as follows: The current vehicle speed v is provided in real time by the vehicle's powertrain system. Automatically generated by the autonomous driving module based on path planning or collision risk assessment; obstacle distance d and safe distance. The deceleration required by the vehicle is calculated in real time by the onboard sensing system or remote monitoring system.

[0020] Based on the above parameter relationships, the target deceleration that will reduce the vehicle to the target speed within a safe distance can be obtained according to the preset deceleration calculation model. Figure 1 This visually demonstrates the basis for calculating the target's deceleration, namely, when... At that time, the vehicle still has a certain buffer distance, and the deceleration can be calculated according to the power-relationship; while when At that time, the vehicle must immediately apply the maximum permissible deceleration to ensure that a collision is avoided.

[0021] Figure 1 The scenario shown only reveals the triggering logic of redundant braking demand. In reality, the braking system of an autonomous vehicle consists of both electric motor braking and electronic parking brake, and its braking availability, response speed, and controllability vary significantly under different operating conditions. Traditional solutions typically only use electric motor braking for service braking. When the performance of the electric motor braking deteriorates or the control link fails, it cannot guarantee that the vehicle will achieve the required deceleration in an emergency.

[0022] Therefore, this application is in Figure 1 Based on the braking demand identification shown, a joint allocation strategy combining road surface adhesion estimation, motor braking availability estimation, and dynamic available deceleration of the electronic parking system is further proposed. By constructing redundant braking performance indicators, intelligent, stable, and adaptive cooperative braking control between the two braking systems is achieved, thereby significantly improving the safety redundancy capability of autonomous vehicles.

[0023] like Figure 2As shown, the redundant braking control system for autonomous vehicles provided in this application embodiment takes the vehicle controller 101 as the core and is connected to the autonomous driving module 102, remote monitoring system 103, driver operation module 104, local data recorder 105, braking actuator (including motor braking system and electronic parking system) and multiple types of safety detection modules through the vehicle communication network, forming a hardware platform that can support target deceleration calculation, road surface adhesion estimation, redundant braking joint distribution and air pressure closed-loop control of electronic parking system.

[0024] The autonomous driving module 102 is used to perform autonomous driving functions such as environmental perception, path planning, and collision risk assessment. When the autonomous driving module 102 determines that the vehicle needs to decelerate or brake, it sends information such as the braking request, the desired target speed, and the distance to obstacles ahead to the vehicle controller 101 via a communication link. The remote monitoring system 103 is used to receive remote control commands issued by the cloud monitoring platform or dispatch center, and sends deceleration or stopping commands to the vehicle controller 101 in unmanned operation, abnormal takeover, or emergency response scenarios. The above two, together with the driver operation module 104, constitute the multi-source input channel for braking requirements in this application. The driver operation module 104 includes manual control devices such as the brake pedal 1041, and can directly send braking requests to the vehicle controller 101 when autonomous driving is disengaged or the driver takes over actively. The vehicle controller 101 integrates the signals from the autonomous driving module 102, the remote monitoring system 103, and the driver operation module 104, and calculates the target deceleration that meets driving safety requirements by combining the current vehicle speed, the distance to obstacles ahead, and the preset safety distance.

[0025] The local data logger 105 is used to record the input and output data of the vehicle controller 101, including various braking request signals, target deceleration, motor braking torque, electronic parking system air pressure, and vehicle longitudinal acceleration, so as to perform playback analysis and safety assessment of the redundant braking control process, and also to provide a traceable data basis for road surface adhesion estimation and online correction of the mapping table in this application.

[0026] Figure 2The system also includes various safety detection modules, such as an advanced defense warning system 1071, ultrasonic radar 1072, anti-collision switch 1073, human body sensing device 1074, and local emergency stop switch 1075. These modules can directly send braking or stopping requests to the vehicle controller 101 when a potential collision risk or intrusion into a dangerous area is detected, serving as a physical safety redundancy channel independent of the autonomous driving algorithm and remote monitoring system 103. The above-mentioned hardware modules work together to provide a complete perception, decision-making, and execution foundation for the redundant braking control method of this application. This enables the autonomous vehicle to reliably track the target deceleration and ensure redundant braking safety through the joint control of the motor braking system and the electronic parking system, even under different road surface adhesion conditions, changes in motor braking capacity, and partial sensor failures.

[0027] In this embodiment, a computer-readable storage medium is provided, on which a computer program for executing the redundant braking control method of this application is stored. The computer-readable storage medium may be a read-only memory (ROM), random access memory (RAM), flash memory, EEPROM, hard disk, solid-state drive, memory card, optical disk, or other medium capable of storing computer instructions.

[0028] In a vehicle control system, the computer program is deployed in the memory of the vehicle control unit (VCU) and loaded and run by the processor inside the VCU.

[0029] like Figure 2 and Figure 3 As shown, the vehicle controller integrates multiple functional modules for specifically executing the redundant braking control method of this application. This method includes a vehicle controller applied to an autonomous vehicle comprising an electric motor braking system and an electronic parking system, the method comprising:

[0030] S1. When the autonomous driving module, collision warning system and / or remote monitoring system detect that the vehicle needs to brake, the current vehicle speed, target vehicle speed, distance to the obstacle ahead and preset safety distance are obtained, and the target deceleration is calculated based on the parameters.

[0031] When the vehicle is operating in autonomous driving mode or remote monitoring mode, the vehicle controller 101 continuously receives braking request signals from the autonomous driving module 102, the forward collision warning system, and the remote monitoring system 103. The autonomous driving module 102 identifies vehicles / obstacles ahead based on onboard sensors, including cameras, millimeter-wave radar, and lidar. If it determines there is a risk of rear-end collision based on the currently planned path, it sends a deceleration or braking request to the vehicle controller 101. This request includes the desired target vehicle speed. And the real-time distance d to the vehicle in front.

[0032] When millimeter-wave radar or a monocular camera detects an obstacle ahead whose distance remains below a warning distance threshold, the advanced defense warning system 1071 actively outputs an emergency braking trigger signal to the vehicle controller 101. At this time, without relying on the path planning results of the autonomous driving module 102, the safety system can directly initiate a forced braking request to avoid a collision.

[0033] In unmanned operation or vehicle takeover scenarios, the remote monitoring center can send remote deceleration or stopping commands to the vehicle controller 101 via wireless communication. For example, when the cloud monitors that the vehicle violates a no-parking zone or that there are dangerous people nearby, the remote monitoring system 103 can directly trigger a redundant braking process.

[0034] Upon receiving a braking request from any of the aforementioned modules, the vehicle controller will simultaneously collect: the current vehicle speed v (provided by the vehicle speed sensor), the distance d of the obstacle ahead provided by the autonomous driving module or ultrasonic radar, and the preset safety distance in the system. The target vehicle speed output by the autonomous driving module, collision warning system, and remote monitoring system. .

[0035] The calculation of the target deceleration in S1 includes:

[0036] Get the current vehicle speed v and the target vehicle speed. Distance d of the obstacle ahead and preset safety distance ;when At that time, the target deceleration is calculated according to the following relationship: ;

[0037] And decelerate the target The braking amplitude is limited to a second preset deceleration threshold, which represents the maximum expected deceleration that the vehicle is allowed to apply under the current operating conditions. In this embodiment, the second preset deceleration threshold is used to limit the maximum deceleration that the system is allowed to apply, so as to ensure the stability and reliability of the braking process. This threshold is determined by comprehensively considering the combined braking capability of the electric motor braking system and the friction braking system, as well as the vehicle test calibration results. On the one hand, the maximum deceleration that the electric motor braking can provide is calculated by using the maximum available braking torque of the motor, the vehicle mass, and the wheel radius. Combined with the friction braking deceleration that the electronic parking system can provide under the maximum output air pressure under the current operating conditions, the maximum comprehensive deceleration that the vehicle can achieve at the hardware level is obtained. On the other hand, the maximum braking capability test of the vehicle is carried out under different loads, different road surfaces, and different braking intensities to obtain the maximum stable deceleration that the vehicle can achieve without sideslip, fishtailing, or triggering braking system overload. Finally, the second preset deceleration threshold of this embodiment is determined by comprehensively considering tire adhesion characteristics, braking system thermal fade capability, and relevant regulatory requirements for braking performance, and by reducing the above theoretical calculation values ​​and test results according to a preset safety factor. For example, for a typical passenger car, this threshold can be set to 6 m / s².

[0038] when When the target deceleration is directly set to the second preset deceleration threshold, a braking process with the maximum permissible deceleration is triggered. When the distance d to the detected obstacle ahead is less than the preset safe distance... When the vehicle enters an emergency avoidance zone, the effective braking response time is extremely short. To avoid further approaching the obstacle and causing potential collision risks, this embodiment directly sets the target deceleration to a second preset deceleration threshold, which is the maximum deceleration that the vehicle can achieve within the limits of hardware capabilities, tire adhesion conditions, and safety stability. This ensures that the braking system can immediately output the strongest braking capacity, triggering the maximum permissible deceleration braking process, and achieving a rapid and effective avoidance maneuver. This method significantly shortens the braking distance, improves safety redundancy under extreme conditions, and avoids delays caused by step-by-step calculations that could affect braking performance.

[0039] In S1, the vehicle controller is further configured to classify braking requirements based on a comparison between the target deceleration and a preset deceleration threshold, including:

[0040] When the target deceleration is greater than or equal to the first preset deceleration threshold, the current braking demand is determined as an emergency braking demand, and in S3, the sum of the motor braking deceleration and the electronic parking deceleration is limited to be no less than the target deceleration, so as to prioritize braking safety.

[0041] When the target deceleration is less than the first preset deceleration threshold, the current braking demand is determined as a general braking demand. In S3, by increasing the penalty weight of the electronic parking deceleration, the motor braking deceleration bears the main part of the target deceleration.

[0042] S2. Obtain the actual longitudinal acceleration of the vehicle, the air pressure of the parking spring chamber of the current electronic parking system, and the braking torque of the motor. Estimate the current total braking force based on the vehicle mass, and obtain the braking force of the electronic parking system after deducting the braking force corresponding to the braking torque of the motor. Based on this, estimate the current road adhesion coefficient and / or equivalent adhesion capability.

[0043] In this embodiment, the vehicle's dynamic parameters in step S2 are acquired in real time through onboard sensors and actuator control units. Specifically, the vehicle's actual longitudinal acceleration is measured by the inertial measurement unit (IMU) or the electronic stability control (ESC) system configured in the vehicle, and transmitted to the vehicle controller via the onboard CAN bus in the form of periodic signals, thereby obtaining real-time longitudinal acceleration data. The parking spring air chamber pressure is collected by a pressure sensor located inside the air chamber of the electronic parking system, and the pressure signal is uploaded to the vehicle controller via the electronic parking controller to characterize the current output capability of the electronic parking system. The motor braking torque is calculated by the motor controller based on the regenerative braking condition and is provided to the vehicle controller via the CAN bus in real time. After acquiring the above three types of data, the vehicle controller calculates the current total braking force using the vehicle mass and the actual longitudinal acceleration, and after deducting the braking force corresponding to the motor braking torque, obtains the estimated value of the actual braking force of the electronic parking system. Then, based on the ratio of this braking force to the vehicle's normal load, it estimates the current road adhesion coefficient and / or equivalent adhesion capability.

[0044] Among them, the vehicle controller first obtains the vehicle mass. and the vehicle's actual longitudinal acceleration The product of the two is used to estimate the total braking force of the vehicle in the longitudinal direction, i.e. At the same time, the vehicle controller obtains the current motor braking torque from the motor controller. Combined with the transmission ratio and the effective radius of the wheels Calculate the braking force generated by the electric motor at the wheels. This braking force is then subtracted from the total braking force to obtain an estimated value of the braking force generated by the electronic parking system. .

[0045] Furthermore, the vehicle controller calculates the load based on the product of the vehicle's mass and gravitational acceleration, or by considering the equivalent normal load after load transfer. The current road surface adhesion coefficient or equivalent adhesion capacity is estimated by using the ratio of total braking force to equivalent normal load. The adhesion coefficient or equivalent adhesion capability is used as a reference parameter for calculating the maximum available deceleration limit of the electronic parking system, and is used for subsequent redundant braking joint allocation.

[0046] Furthermore, in this embodiment, the vehicle controller pre-stores a nominal vehicle mass parameter, which is used to convert the actual longitudinal acceleration of the vehicle into the current total braking force in step S2. The nominal vehicle mass can be obtained through experimental calibration based on the vehicle's curb weight, maximum design gross weight, and typical load conditions. For example, a compromise value between the curb weight and the maximum design gross weight can be taken, or a conservative mass value biased towards full-load conditions can be adopted.

[0047] It should be noted that, since the actual vehicle mass may deviate due to changes in the number of passengers and cargo during actual operation, the vehicle mass in this embodiment is not strictly equal to the vehicle mass at any instant, but is used as an equivalent mass parameter under the current operating conditions. The error introduced can be compensated to a certain extent by subsequent estimation of the road adhesion coefficient and online correction of the electronic parking deceleration air pressure mapping table, so as not to affect the effectiveness and stability of the redundant braking control strategy of this invention.

[0048] In other alternative embodiments, the vehicle controller can also perform online correction of the vehicle mass based on the longitudinal dynamic characteristics of the vehicle during driving or braking. For example, given the driving or braking torque, the current equivalent vehicle mass can be calculated using measured longitudinal acceleration, and this equivalent mass can be used to replace the fixed nominal mass in the calculation of total braking force and adhesion capability through first-order filtering or amplitude limiting update, thereby further improving the accuracy of braking force estimation.

[0049] In other alternative embodiments, considering the significant differences in vehicle mass (especially commercial vehicles) under unloaded, partially loaded, and fully loaded conditions, using only a fixed nominal vehicle mass would lead to distorted estimation of the road adhesion coefficient. Therefore, an online vehicle mass estimation mechanism can be introduced in step S2. Constructing the vehicle's longitudinal dynamics model: ;in The driving force of the vehicle (which can be calculated from the motor torque and transmission ratio); The total vehicle weight to be estimated; It is longitudinal acceleration; air resistance ; For rolling resistance; This is the slope resistance.

[0050] The process for estimating the overall vehicle weight is as follows:

[0051] The first step is data filtering. Select segments where the vehicle is in stable driving and the speed changes gradually as the effective observation window, and remove data from the gear shifting process or large dynamic steering process to reduce noise interference.

[0052] The second step is to... The dynamic equations are transformed into In the form of, The parameters to be identified include quality. Iterative computation is performed using recursive least squares (RLS) with a forgetting factor. RLS includes calculating the prediction error; updating the gain matrix; and updating the parameter estimates. .

[0053] The third step is result smoothing and limiting, which involves applying a moving average filter to the instantaneous quality estimate of the RLS output.

[0054] The final corrected vehicle mass is obtained by applying physical constraints based on the vehicle's curb weight (lower limit) and maximum design gross weight (upper limit).

[0055] As another implementation method, the Kalman filter algorithm can be used to establish a state equation with the vehicle mass as a state variable for observation and updating.

[0056] When the braking logic in step S2 is entered, the vehicle controller calls the latest corrected vehicle mass. Combined with real-time longitudinal acceleration Calculate the current total braking force .

[0057] Through the above-described online mass estimation steps, this embodiment can dynamically compensate for the impact of load changes. Compared with the existing technology that uses a fixed nominal mass, the error of the total braking force calculated by this method can be reduced by more than 15%, thereby making the calculated electronic parking brake system braking force and road adhesion coefficient more accurate. This effectively avoids the problem of excessive EPB braking force distribution causing lock-up or insufficient braking due to incorrect mass estimation.

[0058] In this embodiment, step S2 is executed based on the vehicle having entered the braking process or at least being in a partial braking state, meaning the vehicle controller detects that the electric motor braking system and / or the electronic parking brake system have generated actual braking force output. Only when the vehicle exhibits a genuine longitudinal deceleration response can the current total braking force be accurately inferred from the vehicle's mass, and the contribution of the electric motor braking can be further subtracted to obtain an estimate of the braking force of the electronic parking brake system. Simultaneously, the estimation of the road surface adhesion coefficient or equivalent adhesion capability depends on the vehicle's longitudinal dynamic characteristics under braking conditions; therefore, the above estimation is only performed during braking to ensure the validity and reliability of the parameters.

[0059] S3. When there is redundant braking demand, based on the target deceleration, the current road surface adhesion coefficient and / or equivalent adhesion capability, the upper limit of the maximum available deceleration of the motor braking under the current working condition, and the upper limit of the maximum available deceleration of the electronic parking system under the current working condition, a performance index is constructed with the total deceleration deviation term and the electronic parking deceleration penalty term as the target, and the motor braking deceleration and electronic parking deceleration are jointly allocated to obtain the target motor braking deceleration and the target electronic parking deceleration.

[0060] Redundant braking demand refers to the operating condition where, when a vehicle is performing service braking, the electric motor braking system alone cannot meet the target deceleration requirements, or when a second braking channel (electronic parking system) needs to be activated simultaneously for safety reasons to improve braking reliability.

[0061] Specifically, the vehicle controller determines that there is a redundant braking requirement when any of the following conditions are met: First, the motor braking capacity is insufficient. When the maximum available deceleration corresponding to the maximum available braking torque of the motor calculated by the motor controller based on factors such as the current motor temperature, motor speed, SOC, and inverter limitations is lower than the target deceleration, it indicates that relying solely on motor braking cannot achieve the desired braking effect, and the electronic parking system needs to be activated to participate in braking.

[0062] Secondly, the limited road surface adhesion necessitates dual-channel braking. When the current road surface adhesion coefficient estimated in S2 is lower than a preset threshold, making it impossible for a single braking system to fully utilize the adhesion, an electronic parking system is needed to compensate for the motor braking in order to improve braking force stability and braking distribution margin.

[0063] Thirdly, when S1 determines that there is an emergency braking demand (e.g., the target deceleration is ≥ the first preset deceleration threshold), in order to avoid insufficient braking due to the performance degradation of a single braking channel, the vehicle controller directly activates the redundant braking strategy, so that the electronic parking system and the motor braking system participate in braking in parallel, thereby prioritizing the braking safety of the vehicle.

[0064] Fourthly, when the vehicle controller detects situations such as delayed motor braking response, limited power, temperature protection, or the inverter entering derating mode, it assumes that the reliability of the motor braking channel has decreased and thus activates the electronic parking system in advance to provide redundant braking force.

[0065] Based on the above determination results, when there is redundant braking demand, the vehicle controller will jointly allocate the motor braking deceleration and electronic parking deceleration according to the performance index optimization model in step S3, so that the two form coordinated braking, so as to ensure that the total deceleration is close to the expected target deceleration.

[0066] In this embodiment, the maximum available deceleration limit of the motor braking under the current operating condition is calculated by the vehicle controller based on the maximum available braking torque reported by the motor controller. Specifically, the motor controller queries the pre-calibrated maximum regenerative braking characteristic diagram to obtain the maximum available regenerative braking torque under the current operating condition based on operating parameters such as the current vehicle speed or motor speed, the state of charge (SOC) of the power battery, battery temperature, motor temperature, and inverter operating status. This torque is then limited in conjunction with the maximum allowable charging power of the power battery to obtain the maximum available braking torque of the motor braking. After obtaining the maximum available braking torque, the vehicle controller combines it with the nominal or online estimated vehicle mass. and the equivalent radius of the wheel ,according to The relationship is used to calculate the maximum available deceleration limit of the motor braking under the current operating conditions. This maximum available deceleration limit serves as an upper bound constraint on the motor braking deceleration during the redundant braking joint allocation process, ensuring that the motor braking command does not exceed its safe output capacity under the current operating conditions.

[0067] In other alternative embodiments, the vehicle controller can further constrain the maximum usable deceleration limit of the electric motor braking by combining the road adhesion coefficient estimated in step S2. The deceleration should not exceed the maximum deceleration allowed by the current adhesion conditions, thereby avoiding tire slippage or decreased vehicle stability caused by unilaterally increasing the motor braking deceleration under low adhesion conditions.

[0068] In this embodiment, when the vehicle controller determines that there is a redundant braking requirement, it jointly allocates the motor braking deceleration and the electronic parking brake deceleration. To this end, the vehicle controller constructs a performance index function of the following form based on the target deceleration, the current road surface adhesion coefficient, and the maximum available deceleration limit of the motor braking and electronic parking brake systems under the current operating conditions: ,in, The target deceleration obtained in step S1, and These are the decelerations to be allocated to the electric braking system and the electronic parking brake system at the current moment, respectively. This is the weighting coefficient for the total deceleration deviation term, used to constrain the actual total deceleration to be as close as possible to the target deceleration. This is a weighting coefficient for the electronic parking brake deceleration penalty term, used to limit the use of the electronic parking brake system under normal braking conditions.

[0069] In addition to the above performance indicators, the vehicle controller also imposes the following constraints: .

[0070] in, This is the upper limit of the maximum usable deceleration calculated based on the current braking capacity of the motor. This represents the maximum usable deceleration limit obtained by combining the current road surface adhesion coefficient and the air pressure capability of the electronic parking brake system. The vehicle controller obtains the optimal solution that minimizes the performance index J by solving the constrained performance index minimization problem described above. These values ​​are then used as the target motor braking deceleration and the target electronic parking deceleration for subsequent braking execution control.

[0071] Furthermore, under normal braking conditions, the weighting coefficient of the electronic parking deceleration penalty term can be adjusted. Set to a larger value to reduce the involvement of the electronic parking system and allow the electric motor brake to bear the main deceleration; however, in emergency braking conditions, it can be appropriately reduced. Alternatively, the weight of the total deceleration deviation term can be increased to prioritize ensuring that the actual total deceleration is not lower than the target deceleration, thereby prioritizing braking safety.

[0072] In a preferred embodiment of this invention, a weighting coefficient is used to adjust the degree of participation of the electronic parking system. A calculation method that dynamically changes with braking demand is adopted. Specifically, the vehicle controller first calculates the target deceleration obtained in step S1. And the upper limit of the maximum available deceleration for motor braking calculated based on the current operating conditions. Construct braking demand indicators ,in, This indicates that the input value is limited within the range of 0 to 1. Subsequently, the vehicle controller operates within the preset upper and lower limits of the weighting coefficients. and Between Performing linear interpolation, we obtain: .

[0073] Therefore, when the braking demand is small and the deceleration requirement can be met by relying solely on motor braking, the severity index S is small, and the weighting coefficient is small. Approaching the upper limit The optimization process significantly suppresses the distribution of electronic parking deceleration, causing the electric motor brake to bear the main deceleration; however, when the target deceleration approaches the upper limit of the electric motor braking capacity, the severity index S approaches 1, and the weighting coefficient... Reduce to near This reduces the suppression of electronic parking deceleration, allowing the electronic parking system to obtain a larger share of deceleration in the optimized allocation, ensuring that the total deceleration can meet current safety requirements.

[0074] In other alternative embodiments, the weighting coefficient It can also be corrected by combining the road adhesion coefficient estimated in step S2, for example, by appropriately reducing it under low adhesion conditions. This increases the proportion of the electronic parking brake system participating in redundant braking, thereby further ensuring the braking safety of the entire vehicle when the braking capacity of the electric motor is limited.

[0075] S4. Based on the electronic parking deceleration air pressure mapping table containing basic calibration data and online correction values, the target electronic parking deceleration is converted into the corresponding target air pressure of the parking spring chamber. Combined with the real-time air pressure feedback collected by the air pressure sensor, the air pressure of the parking spring chamber is controlled in a closed loop to make the actual air pressure approach the target air pressure, thereby achieving redundant driving braking close to the target deceleration together with the target motor braking deceleration.

[0076] In this embodiment, to enable the electronic parking brake system to achieve stable, controllable, and adaptive braking force output based on the current operating conditions during redundant braking, a set of electronic parking brake deceleration air pressure mapping tables is pre-built inside the vehicle controller, and the mapping relationship is adaptively updated during operation in conjunction with online correction values. This mapping table consists of two parts: basic calibration data and online correction values, and its implementation is as follows.

[0077] First, during the vehicle testing phase, the deceleration of the vehicle under different parking spring chamber pressures when braking was applied independently by the electronic parking system was collected. The tests were conducted under standard load, standard tires, and standard road conditions. By statistically analyzing the braking response data at multiple different air pressure points, a discrete mapping relationship between the "target electronic parking deceleration and the base air pressure" was established.

[0078] Target electronic parking deceleration (m / s²) Corresponding base pressure (kPa) illustrate 3 100 With lower air pressure, the spring is fully released, resulting in greater braking force. 2.75 120 2.5 140 2.25 160 2 185 1.75 210 1.5 235 1.25 260 1 285 0.75 310 0.5 335 0.25 360 The air pressure is high, the spring has just begun to extend, providing slight braking.

[0079] (Table 1, Electronic Parking Deceleration Air Pressure Mapping Table, indicates the basic air pressure value required for the parking spring air chamber to achieve the corresponding electronic parking deceleration.)

[0080] In this embodiment, the vehicle controller uses piecewise linear interpolation or other interpolation algorithms to support rapid lookup of the basic air pressure for any input target electronic parking deceleration.

[0081] The basic calibration data obtained in the above manner serves as the benchmark for electronic parking brake air pressure control, and is used to provide braking capacity curves under standard operating conditions.

[0082] Based on the aforementioned mapping table, to address braking response deviations caused by factors such as vehicle load variations, tire wear, temperature changes, and road surface adhesion coefficient variations in actual operating conditions, this embodiment configures an independent online correction amount ΔP(i) for each basic air pressure level in the mapping table. During redundant braking execution, the vehicle controller adjusts the braking response based on the actual longitudinal acceleration of the vehicle. And the actual contribution deceleration of the electronic parking system is calculated using the motor contribution deceleration estimated through motor braking. and the target electronic parking deceleration The deceleration error is obtained by comparison: The vehicle controller then updates the online correction amount ΔP(i) for the corresponding gear according to the preset learning rate ki, where ΔP(i) = ΔP(i) + ki·e, typically ki ∈ [0.01, 0.05].

[0083] Furthermore, to avoid excessive correction causing a sudden increase or decrease in braking force, this embodiment sets an amplitude constraint on the correction amount: ΔPmin≤ΔP(i)≤ΔPmax.

[0084] When the target electronic parking deceleration is obtained in step S3 Subsequently, the vehicle controller in this embodiment performs the following calculation process based on the above mapping table:

[0085] S41, obtain the corresponding baseline air pressure by finding the baseline calibration data and interpolating it.

[0086] S42, calculate the corresponding online correction amount ΔP(i) based on the interval where the deceleration is located.

[0087] S43, the target air pressure of the parking spring chamber is obtained by superimposing the base air pressure and the online correction. .

[0088] The vehicle controller controls the opening and closing of the solenoid valve of the electronic parking system based on the deviation between the target air pressure and the air pressure fed back by the sensor, so that the actual air pressure gradually approaches the target air pressure, thereby outputting braking force that matches the target electronic parking deceleration.

[0089] In this embodiment, while the vehicle controller performs the closed-loop air pressure control of the electronic parking system in step S4, in order to ensure that the braking effect of the vehicle under dynamic conditions always meets the target deceleration requirements, a mechanical feedback mechanism is further introduced to monitor the braking effect and execute a gain adjustment strategy.

[0090] First, the vehicle controller is based on the vehicle's longitudinal acceleration sensor within a preset monitoring time window. Continuous collection of actual longitudinal acceleration of vehicles The acceleration was then subjected to a moving average filter to eliminate the effects of road excitation, vibration, and sensor noise, thus obtaining the actual total deceleration. Subsequently, the vehicle controller compares the actual total deceleration with the target total deceleration obtained from the optimized allocation in step S3. By comparing the values, the deceleration deviation is determined: .

[0091] When detected The deceleration remains below the target deceleration multiplied by a preset scaling factor throughout the entire monitoring window. (For example, 0.9), which satisfies: If the braking effect is insufficient, the vehicle controller determines that the electronic parking brake system is not effective enough. To compensate for the effects of response lag, air pressure error, or decreased adhesion during actual braking, the vehicle controller appropriately increases the braking force contribution of the electronic parking brake system. Specifically, this includes increasing the target electronic parking brake deceleration obtained in step S3. Alternatively, without changing the target deceleration, the corresponding base air pressure or corrected air pressure in the mapping table can be appropriately reduced to improve the actual braking effect.

[0092] To prevent sudden increases in braking force from causing longitudinal impact or exceeding the adhesion utilization limit, the vehicle controller incorporates step limits for gain adjustments. For example, within a single adjustment cycle, the target electronic parking deceleration is only allowed to increase by no more than 0.1 m / s², or the target air pressure is allowed to adjust by no more than 5 kPa. The limiting module ensures that the output of the electronic parking system does not exceed the maximum available deceleration limit allowed by the current road conditions.

[0093] Through the aforementioned monitoring and gain adjustment mechanism, this embodiment can dynamically adjust the action ratio of the electronic parking system in real time according to the actual response of the vehicle during braking execution, making it closer to the theoretical deceleration distribution, effectively improving the stability and reliability of redundant braking control, and ensuring that the vehicle can still achieve a braking effect close to the target deceleration under varying road surface and load conditions.

[0094] When the actual total deceleration is consistently higher than the target deceleration multiplied by another preset proportional coefficient, the target electronic parking deceleration is reduced and / or the corresponding target air pressure is increased, thereby reducing the braking effect of the electronic parking system and avoiding excessive deceleration.

[0095] In this embodiment, to ensure that the redundant braking control system still has controllable and predictable braking performance when key sensors malfunction or road surface adhesion estimation is inaccurate, the vehicle controller further sets up an algorithm degradation strategy. When any key sensor used to obtain actual longitudinal acceleration, electronic parking air pressure, or motor braking torque is detected to be faulty, or when the road surface adhesion coefficient calculated in step S2 exceeds a preset confidence range (e.g., below 0.05 or above 0.95), the vehicle controller immediately activates the degradation mode and switches to the redundant braking control logic in a safe state.

[0096] Specifically, the vehicle controller first pauses the online correction process of the electronic parking deceleration air pressure mapping table described in step S4. To avoid control deviations caused by the continuous accumulation of correction values ​​due to fault data, the controller freezes the current online correction value or resets it to zero, bringing the air pressure control back to the basic calibration level. Simultaneously, all deceleration error information from the moment of sensor failure is automatically masked and no longer participates in the mapping table learning process.

[0097] Subsequently, the vehicle controller replaces the performance index optimization allocation strategy used in step S3 with a fixed proportion or fixed priority braking force allocation method. For example, under default conditions, 70% of the target deceleration can be fixedly allocated to the electric motor braking, with the electronic parking system handling the remaining 30%; or, if there are fault trends such as abnormal noise or torque fluctuations on the motor side, the electronic parking system will be given priority to handle a higher proportion of deceleration. This fixed allocation strategy avoids drastic deceleration fluctuations caused by optimization calculations relying on inaccurate data.

[0098] Furthermore, in degraded mode, the vehicle controller strictly limits the maximum output air pressure and maximum permissible deceleration of the electronic parking brake system to prevent over-braking due to misjudgment in the event of sensor failure. Simultaneously, all deceleration commands employ more conservative slope limits (e.g., maximum rate of change not exceeding 0.5 m / s² / s) to improve vehicle longitudinal stability.

[0099] After the fault is cleared, the vehicle controller confirms that all sensors have returned to normal operation through a sensor self-test program, and verifies that the road adhesion estimation has recovered to a reliable range through multiple samplings. After the fault signal disappears, the vehicle must operate continuously for at least T seconds (e.g., 3 seconds) and the braking state must be in a non-operating or low-load state before switching back to normal mode. After confirming normal operation, the controller smoothly switches the braking force distribution strategy from a fixed distribution mode back to a performance-optimized distribution algorithm, while simultaneously unfreezing the online correction amount, allowing the electronic parking deceleration pressure mapping table to resume adaptive updates. During this process, the controller employs a transition buffer strategy, such as setting a 2-3 second gradual recovery phase, to avoid sudden changes in the control law that could cause longitudinal impacts on the vehicle.

[0100] Through the aforementioned algorithm degradation mechanism, this embodiment can maintain the basic controllability and safety of the braking system in the event of sensor malfunction, data distortion, or unreliable adhesion estimation. At the same time, it can naturally recover to the optimized control mode after the fault is cleared, thereby achieving fault tolerance and robustness of the system.

[0101] In this embodiment, in order to enable the electronic parking deceleration air pressure mapping table to maintain stable and rapid convergence characteristics under different road conditions such as dry, wet, snowy, and icy conditions, the vehicle controller further manages the online correction amount in the mapping table using a strategy of "independent maintenance according to road surface adhesion interval".

[0102] Specifically, the vehicle controller first divides the range of the road adhesion coefficient μ into several non-overlapping adhesion intervals based on typical road adhesion conditions. For example, μ can be divided into:

[0103] μ∈[0.0,0.2] (low snow and ice adhesion interval)

[0104] μ∈(0.2,0.4] (low-to-medium adhesion range)

[0105] μ∈(0.4,0.7] (attached interval in general roads)

[0106] μ∈(0.7,1.0] (dry, high adhesion range).

[0107] For each attachment range, the vehicle controller maintains an independent set of "electronic parking deceleration air pressure mapping table online correction values" so that the mapping relationship under different road conditions can be learned independently and without interference.

[0108] When the vehicle performs redundant braking, the vehicle controller first determines the current adhesion interval of the vehicle based on the road adhesion coefficient estimation result μest in step S2. If μest falls into a specific interval, a set of online correction values ​​corresponding to that interval is loaded from the memory and used together with the pre-calibrated basic mapping table to calculate the target air pressure corresponding to the current target electronic parking deceleration.

[0109] Meanwhile, the vehicle controller only updates the online correction amount corresponding to the attachment interval: after calculating the actual contribution deceleration of the electronic parking system and comparing it with the target electronic parking deceleration, the deceleration error is accumulated into the online correction amount of the current interval according to the preset learning rate, and the upper and lower limits of the correction amount are constrained in amplitude, thereby realizing the adaptive adjustment of the mapping relationship for the road surface attachment interval.

[0110] When changes in the vehicle's driving environment cause μest to switch from the first attachment range to the second attachment range, the vehicle controller will automatically switch to using a set of online correction values ​​corresponding to the second attachment range and pause updating the data for the first attachment range. Through this strategy, the electronic parking brake response curves under different road conditions can learn independently and gradually converge, avoiding the problem of a single mapping being repeatedly "incorrectly pulled" under multiple conditions, leading to convergence difficulties. This ensures that the mapping table maintains accuracy and stability under various conditions such as dry roads, wet roads, and icy roads, providing reliable input for the air pressure control in step S4.

[0111] In summary, the redundant braking control method for autonomous vehicles proposed in this invention achieves stable and reliable braking control under complex operating conditions, sensor malfunctions, and fluctuations in braking system performance by dynamically calculating the target deceleration, estimating the road surface adhesion coefficient in real time, and optimizing the allocation of motor braking and electronic parking brake. By introducing an electronic parking deceleration pressure mapping table and its online adaptive correction mechanism, the electronic parking system can automatically adjust its output according to different adhesion conditions and load changes, thereby improving the response accuracy and consistency of redundant braking. Furthermore, through the comprehensive design of braking effect monitoring, fault diagnosis, and algorithm degradation strategies, this invention further enhances the safety redundancy capability of autonomous vehicles in extreme scenarios. It should be noted that those skilled in the art can adjust the specific structure or parameters without departing from the core concept of this invention, and these adjustments should be considered within the scope of protection of this invention.

Claims

1. A redundant braking control method for an autonomous vehicle, applied to the vehicle controller of an autonomous vehicle including an electric braking system and an electronic parking system, characterized in that, The method includes: S1. When the autonomous driving module, collision warning system and / or remote monitoring system detect that the vehicle braking requirement is triggered, the current vehicle speed, target vehicle speed, distance to the obstacle ahead and preset safety distance parameters are obtained, and the target deceleration is calculated based on the parameters. S2. Obtain the actual longitudinal acceleration of the vehicle, the air pressure of the parking spring chamber of the current electronic parking system, and the braking torque of the motor. Estimate the current total braking force based on the vehicle mass, and obtain the braking force of the electronic parking system after deducting the braking force corresponding to the braking torque of the motor. Based on this, estimate the current road adhesion coefficient and / or equivalent adhesion capability. S3. When there is redundant braking demand, based on the target deceleration, the current road surface adhesion coefficient and / or equivalent adhesion capability, the maximum available deceleration limit of the motor braking under the current working condition and the maximum available deceleration limit of the electronic parking system under the current working condition, a performance index function with the total deceleration deviation term and the electronic parking deceleration penalty term as the target is constructed, and the motor braking deceleration and electronic parking deceleration are jointly allocated to obtain the target motor braking deceleration and the target electronic parking deceleration. The performance metric function is constructed as follows: , The target deceleration obtained in step S1, and These are the decelerations to be allocated to the electric braking system and the electronic parking brake system at the current moment, respectively. This is the weighting coefficient for the total deceleration deviation term, used to constrain the actual total deceleration to be as close as possible to the target deceleration. The weighting coefficient for the electronic parking deceleration penalty term is used to limit the use of the electronic parking system under normal braking conditions. As braking demand changes dynamically, based on the target deceleration And the upper limit of the maximum available deceleration for motor braking calculated based on the current operating conditions. Construct braking demand indicators ,in This indicates that the input value is limited within the range of 0 to 1; subsequently, the vehicle controller operates within the preset upper and lower limits of the weighting coefficients. and Between Perform linear interpolation to obtain ; S4. Based on the electronic parking deceleration air pressure mapping table containing basic calibration data and online correction values, the target electronic parking deceleration is converted into the corresponding target air pressure of the parking spring chamber. Combined with the real-time air pressure feedback collected by the air pressure sensor, the air pressure of the parking spring chamber is controlled in a closed loop to make the actual air pressure approach the target air pressure, thereby achieving redundant driving braking close to the target deceleration together with the target motor braking deceleration. The online correction value is obtained by estimating the actual deceleration contributed by the electronic parking system based on the actual longitudinal acceleration of the vehicle and the motor braking deceleration, and comparing it with the target electronic parking deceleration to obtain the deceleration error. The deceleration error is accumulated into the online correction amount corresponding to the current target electronic parking deceleration according to a preset learning rate, and the amplitude of the online correction amount is constrained.

2. The redundant braking control method according to claim 1, characterized in that, Step S2 specifically includes: The current total braking force is estimated by multiplying the vehicle mass by the actual longitudinal acceleration. The braking force generated by the electric motor braking is estimated based on the electric motor braking torque and the wheel radius, and then subtracted from the current total braking force to obtain an estimated value of the braking force generated by the electronic parking system; The current road surface adhesion coefficient and / or equivalent adhesion capacity are estimated using the ratio of the current total braking force to the vehicle's equivalent normal load, and the adhesion coefficient and / or equivalent adhesion capacity are used as a reference for calculating the upper limit of the maximum available deceleration of the electronic parking system.

3. The redundant braking control method according to claim 2, characterized in that, The electronic parking deceleration pressure mapping table in S4 includes a pre-calibrated mapping relationship between multiple target electronic parking decelerations and their corresponding base pressures, as well as an online correction amount corresponding to each base pressure.

4. The redundant braking control method according to claim 3, characterized in that, In S1, the vehicle controller is further configured to classify braking requirements based on a comparison between the target deceleration and a preset deceleration threshold, including: When the target deceleration is greater than or equal to the first preset deceleration threshold, the current braking demand is determined as an emergency braking demand, and in S3, the sum of the motor braking deceleration and the electronic parking deceleration is limited to not be less than the target deceleration. When the target deceleration is less than the first preset deceleration threshold, the current braking demand is determined as a general braking demand, and in S3, the penalty weight for the electronic parking deceleration is increased.

5. The redundant braking control method according to claim 4, characterized in that, The calculation of the target deceleration in S1 includes: Get the current vehicle speed v and the target vehicle speed. Distance d of the obstacle ahead and preset safety distance ; when At that time, the target deceleration is calculated according to the following relationship: ; And decelerate the target Limit the amplitude to ensure it does not exceed the second preset deceleration threshold. when When the target deceleration is set to the second preset deceleration threshold, the braking process with the maximum permissible deceleration is triggered.

6. The redundant braking control method according to any one of claims 1 to 5, characterized in that, While executing S4, the vehicle controller also monitors braking performance and adjusts gain, including: The actual longitudinal acceleration is collected within a preset monitoring time window, the actual total deceleration is calculated, and it is compared with the target deceleration. When the actual total deceleration is continuously lower than the target deceleration multiplied by a preset proportional coefficient within the monitoring time window, the target electronic parking deceleration is increased and / or the corresponding target air pressure is decreased. When the actual total deceleration is consistently higher than the target deceleration multiplied by another preset proportional coefficient, the target electronic parking deceleration is reduced and / or the corresponding target air pressure is increased.

7. The redundant braking control method according to any one of claims 1 to 5, characterized in that, The vehicle controller is also configured to perform algorithm degradation when it detects a malfunction in a sensor used to obtain actual longitudinal acceleration, electronic parking pressure, or motor braking torque, and / or when the road adhesion coefficient estimation result exceeds a preset confidence interval, including: Suspend the online correction of the electronic parking deceleration barometric pressure mapping meter, and freeze the online correction amount at the current value or reset it to zero; The allocation of the motor braking deceleration and electronic parking deceleration is switched from performance index-optimized allocation to an allocation strategy based on a preset fixed ratio or priority. After the fault is cleared and the sensors are restored to normal, the online correction of the performance index optimization allocation and the electronic parking deceleration pressure mapping table is reactivated.

8. The redundant braking control method according to claim 3 or claim 4, characterized in that, The online correction amount is maintained independently according to the interval to which the current road surface adhesion coefficient belongs, specifically including: The road surface adhesion coefficient range is divided into several adhesion intervals, and each adhesion interval corresponds to an online correction amount of a set of electronic parking deceleration air pressure mapping tables; When the estimated road surface adhesion coefficient is located in different adhesion intervals, the online correction amount corresponding to the adhesion interval is selected to participate in the calculation of the target air pressure, and only the online correction amount corresponding to the adhesion interval is updated; When the road surface adhesion coefficient switches from the first adhesion range to the second adhesion range, the vehicle controller automatically switches to use the online correction amount corresponding to the second adhesion range to adapt the electronic parking deceleration air pressure mapping relationship under different road conditions.

9. The redundant braking control method according to claim 1, characterized in that, The vehicle mass mentioned in step S2 is the corrected vehicle mass obtained through online estimation. The acquisition method includes: collecting vehicle driving force, longitudinal acceleration and vehicle speed signals under non-braking driving conditions; and estimating the vehicle mass in real time using the recursive least squares method based on the vehicle longitudinal dynamics model.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the redundant braking control method as described in claim 1.

Images