Off-highway dump truck multi-power cooperative energy management method and system and dump truck

Abstract

The application discloses a kind of off-highway dump truck multi-power collaborative energy management method, system and dump truck, and existing off-highway dump truck multi-power system energy distribution is unreasonable, control complexity is high, cannot give consideration to power safety and vehicle stable operation, there are optimization solution unreliable, online control response slow and other technical problems.For solving this problem, the application first constructs the model of multi-power system for model predictive control and sets the operating constraint condition, then constructs the optimization problem based on the model and constraint, offline rolling optimization solution forms sample data set, extracts the mapping relationship of power state deviation and torque distribution ratio and solidifies as threshold rule table, real-time acquisition related parameters when vehicle operation, query rule table to obtain target torque distribution ratio and output instruction execution.This method is mainly used for the collaborative energy management of off-highway dump truck multi-power system, realizes the reasonable distribution of multi-power torque, and guarantees the stable operation of vehicle.

Description

Technical Field

[0001] This invention relates to the field of energy management technology for mining trucks. More specifically, this invention relates to a multi-power source collaborative energy management method, system, and dump truck for off-highway dump trucks. Background Technology

[0002] Currently, the main power system architectures used in new energy off-highway dump trucks include series (range-extended) hybrid systems and pure electric power systems. The difference between the two is that the series hybrid system uses a fossil fuel-driven internal combustion engine as a supplementary energy source for the battery, and a generator continuously provides power to the battery, thereby extending the vehicle's range and reducing dependence on charging conditions. In contrast, the pure electric system relies entirely on the battery as its sole energy source, supplemented by external charging equipment. Its power output and range are entirely determined by the battery capacity, discharge performance, and charging conditions.

[0003] Therefore, for pure electric products, the more battery packs there are at the same energy density, the more energy they provide, the longer they can operate, and the higher the transportation and production efficiency. However, under current technology, the larger the battery pack consisting of individual cells connected in series and parallel, the lower its internal insulation performance, leading to increased safety risks for the power battery pack. Therefore, the overall power battery pack cannot be increased indefinitely. However, the usage scenarios of large-tonnage off-highway dump trucks dictate their demand for high traction / braking force. To provide the necessary energy for the operation of the entire machine, the architecture can be designed to isolate multiple battery systems from each other, achieving safe redundancy and coordinated power management across multiple battery systems.

[0004] For off-highway dump trucks employing multiple power systems, it is crucial to maintain balanced power consumption across these systems during operation. Failure to do so can lead to inconsistent drive power on both sides, resulting in vehicle dynamics deviations, reduced regenerative braking capability, accelerated battery life degradation, and even single-side drive system failure causing vehicle shutdown. This severely impacts overall vehicle performance and fleet operational efficiency. Therefore, a collaborative energy management method for multiple power sources in off-highway dump trucks is urgently needed. Summary of the Invention

[0005] Another objective of this invention is to provide a multi-power source collaborative energy management method, system, and dump truck for off-highway dump trucks.

[0006] To achieve these objectives and other advantages according to the present invention, a multi-power source coordinated energy management method for off-highway dump trucks is provided, comprising: Step 1: Construct a multi-power system model for model predictive control and set the operating constraints of the multi-power system. The multi-power system model includes the power state model of each power system, the power-torque relationship model between each power system and the corresponding drive motor, and the vehicle traction and braking demand model. The vehicle traction and braking demand model is used to characterize the torque demand changes under the preset route. The operating constraints include power boundary constraints of the power system, torque boundary constraints of the drive motor, electrical safety boundary constraints, and tire adhesion and braking stability constraints. Step 2: Based on the multi-power system model and operating constraints, construct the model predictive control optimization problem of the multi-power system in the prediction time domain; in offline mode, perform rolling optimization to solve the optimization problem based on preset typical operating conditions, obtain the optimal torque distribution sequence in the prediction time domain, and record the torque distribution value, power status, operating mode and constraint activation status at each time in the optimal torque distribution sequence to form a sample dataset. Step 3: Perform statistical analysis on the sample dataset, extract the mapping relationship between the power status value of each power system and the preset reference value and the torque distribution ratio, and solidify the mapping relationship into a threshold rule table indexed by the power status deviation range and the operating mode. Step 4: When the vehicle is running, the vehicle controller collects the vehicle torque demand, the current power status of each power system and the current operating mode in real time. It calculates the power status deviation of each power system based on the current power status, and queries the threshold rule table based on the power status deviation and the current operating mode to obtain the target torque distribution ratio of each power system. It calculates the target torque of each drive motor based on the target torque distribution ratio and the vehicle torque demand, and outputs the target torque command to each motor controller for execution.

[0007] Preferably, in step one, the power boundary constraints of the power system include the maximum allowable discharge power constraint and the maximum allowable recharge power constraint; The torque boundary constraints of the drive motor include the maximum output torque constraint and the maximum braking torque constraint; The electrical safety boundary constraints include the upper and lower limits of electrical state safety constraints for each power system. The constraints on tire adhesion and braking stability are determined based on the tire-road adhesion coefficient, tire vertical force, transmission ratio, and transmission efficiency.

[0008] Preferably, in step two, the objective function of the model predictive control optimization problem includes a state of charge offset term, a battery degradation penalty term, a torque smoothing term, and a slack variable penalty term. Among them, the state of charge offset term is used to minimize the deviation between the state of charge of each power system and the preset reference value, the battery degradation penalty term approximates the degree of battery degradation with the square of the current, the torque smoothing term is used to suppress the torque change rate, and the slack variable penalty term is used to ensure the solvability of the optimization problem when there are constraint conflicts.

[0009] Preferably, the constraints of the model predictive control optimization problem also include: the relationship between battery power and motor torque, the torque change rate constraint, and control variable relaxation constraints; the control variable relaxation constraints, by introducing relaxation variables, allow the total torque demand to be relaxed when there is a constraint conflict, so as to ensure that there is a feasible solution to the optimization problem.

[0010] Preferably, in step three, extracting the mapping relationship between the power status deviation and the torque distribution ratio specifically includes: taking the deviation between the power status of each power system and the preset reference value as the main feature variable, grouping the sample data according to the operating mode and constraint activation state, dividing the power status deviation within each group into intervals, statistically analyzing the distribution characteristics of the torque distribution ratio within each interval, processing the statistical results to obtain the distribution curve, and solidifying the distribution curve in the form of a threshold rule table for online control.

[0011] Preferably, in step four, after calculating the target torque of each drive motor, a real-time constraint verification and trimming step is also included: The vehicle controller applies power boundary constraints of the power system, torque boundary constraints of the drive motor, torque change rate constraints, and tire adhesion and braking stability constraints to the calculated target torque. When any constraint is triggered, the torque of each branch is limited or redistributed according to preset priority to ensure that the final output torque command meets the system safety boundary.

[0012] Preferably, in step four, when an abnormal state such as abnormal data acquisition status, power system derating, or threshold rule mismatch is detected, the vehicle controller switches to the equal distribution mode or the preset safe distribution mode to ensure the stable operation of the vehicle under abnormal conditions.

[0013] The present invention also provides a multi-power cooperative energy management system for off-highway dump trucks, comprising: Multiple independent power supply systems, each connected to a corresponding drive motor; The vehicle controller is communicatively connected to the motor controllers of the multiple power systems and each drive motor, and is used to execute the multi-power coordinated energy management method for off-highway dump trucks.

[0014] Preferably, it also includes: The real-time constraint verification module is used to perform real-time constraint verification and trimming of the target torque after calculating the target torque of each drive motor. The exception handling module is used to automatically switch to the equal distribution mode or the preset safe distribution mode to distribute torque when an abnormal state is detected.

[0015] The present invention also provides an off-highway dump truck, including the aforementioned off-highway dump truck multi-power source collaborative energy management system.

[0016] The present invention has at least the following beneficial effects: First, by introducing model predictive control under typical routes / operating conditions, this invention simultaneously considers constraints and objectives such as power balance, energy utilization efficiency, lifetime decay suppression, and torque smoothness in the prediction time domain. This allows for the acquisition of a better torque distribution sequence for multi-power supply systems under multi-moment coupling conditions, thereby significantly improving the overall efficiency of multi-power supply collaborative output / recovery and long-term power consistency, and reducing long-term SOC drift and the phenomenon of "single power supply overload".

[0017] Secondly, by performing statistical analysis and feature extraction on the optimal solution of MPC, this invention solidifies the "state-allocation" relationship into a mapping rule in the form of threshold / segmentation / lookup table, so that online control can directly output the target torque allocation result without running an optimization solver, thereby significantly reducing the amount of online calculation and real-time pressure, and improving the deployability and determinism of the control strategy on automotive-grade controllers.

[0018] Third, based on the threshold rule output, this invention further applies power constraints, adhesion constraints, and torque change rate constraints to achieve real-time protection of battery power boundaries, regeneration capacity, and road adhesion limits, avoiding overcurrent, overcharging / over-discharging, drive / braking instability, and torque shocks caused by sudden changes in operating conditions or model deviations, thereby improving the overall vehicle's operational stability and safety redundancy.

[0019] Fourth, the threshold parameters of this invention can be updated online based on vehicle operating data, enabling the rules to adapt and evolve with battery degradation, temperature changes, and load changes, maintaining the consistency of control performance; at the same time, it automatically degrades to an equal distribution or preset safe distribution mode under abnormal conditions, ensuring that executable torque commands can still be output under conditions such as sensor abnormalities, constraint conflicts, or single branch derating, thereby enhancing the system's fault tolerance and fault controllability.

[0020] In summary, this solution, through a technical chain of "offline MPC optimization—rule extraction and solidification—online constraint trimming—parameter adaptation and degradation protection," achieves near-optimal multi-power supply torque collaborative allocation without increasing the complexity of online solution. It balances efficiency, lifespan, safety, and real-time performance, and can improve the system's engineering feasibility, stability, and robustness.

[0021] Other advantages, objectives and features of the present invention will become apparent in part from the following description, and in part from those skilled in the art through study and practice of the invention. Attached Figure Description

[0022] Figure 1 This is a flowchart illustrating the process of obtaining the target torque during vehicle operation, as described in one of the technical solutions of the present invention. Detailed Implementation

[0023] The present invention will be further described in detail below with reference to embodiments, so that those skilled in the art can implement it based on the description.

[0024] According to one embodiment of the present invention, such as Figure 1 As shown, the multi-power source collaborative energy management method for off-highway dump trucks includes: Step 1: Construct a multi-power system model for model predictive control (MPC) and set the operating constraints of the multi-power system. The multi-power system model includes the state of charge model of each power system, the power-torque relationship model between each power system and the corresponding drive motor, and the vehicle traction and braking demand model. The vehicle traction and braking demand model is used to characterize the torque demand changes under the preset route. The operating constraints include power boundary constraints of the power system, torque boundary constraints of the drive motor, electrical safety boundary constraints, and tire adhesion and braking stability constraints. Step 2: Based on the multi-power system model and operating constraints, construct the model predictive control optimization problem of the multi-power system in the prediction time domain; in offline mode, perform rolling optimization to solve the optimization problem based on preset typical operating conditions, obtain the optimal torque distribution sequence in the prediction time domain, and record the torque distribution value, power status, operating mode and constraint activation status at each time in the optimal torque distribution sequence to form a sample dataset. Step 3: Perform statistical analysis on the sample dataset, extract the mapping relationship between the power status value of each power system and the preset reference value and the torque distribution ratio, and solidify the mapping relationship into a threshold rule table indexed by the power status deviation range and the operating mode. Step 4: When the vehicle is running, the vehicle controller collects the vehicle torque demand, the current power status of each power system and the current operating mode in real time. It calculates the power status deviation of each power system based on the current power status, and queries the threshold rule table based on the power status deviation and the current operating mode to obtain the target torque distribution ratio of each power system. It calculates the target torque of each drive motor based on the target torque distribution ratio and the vehicle torque demand, and outputs the target torque command to each motor controller for execution.

[0025] In the above technical solution, a system model for model predictive control is established. Discrete-time indexing is defined. k This indicates the current control period, with a sampling period of Δ. t The length of the prediction domain is N p There are N independent power supply systems in total. For the first... i The power supply-drive subsystem, in the control cycle k The status at any given time includes: battery status SOCIETY i (k) Current ,Voltage motor speed Motor torque The established model includes: The state-of-charge model of each power supply system is used to describe the trend of energy change during the output or recovery of energy. The state-of-charge model is established based on the power integral relationship, as shown in equation (1): (1) in, SOCIETY i For the first i The power status of the power supply, △ t The sampling period is E i,nom Rated power, P b,i Let j be the battery power, and j be the step index in the prediction time domain, j=0,1,…,N p -1 indicates the time from the current moment k The future j step; The power and torque relationship models between each drive motor and the power supply system are used to describe the energy consumption or recovery characteristics of the motor at different speeds and operating modes, as shown in equations (2) and (3): (2) (3) Wherein, equation (2) represents the discharge mode, and equation (3) represents the charging mode. For discharge efficiency, For charging efficiency; A vehicle traction and braking demand model is used to characterize the changes in torque demand under a preset route. Based on preset typical operating conditions (such as speed curves and gradient information), the model calculates the sequence of vehicle torque demand at future time points through longitudinal vehicle dynamics calculations. The positive values ​​represent traction demand, and the negative values ​​represent braking demand. The specific form of this model can be established based on vehicle parameters (such as vehicle mass, frontal area, rolling resistance coefficient, etc.) and operating condition information, but this patent does not limit its specific mathematical form, only requiring that it can provide a torque demand sequence in the predicted time domain as input to the MPC optimization problem.

[0026] The constraints in step one include: Power boundary constraints of power supply systems; Torque boundary constraints of drive motor: Maximum output / braking torque limit of motor; The electrical safety boundary constraint is shown in equation (4): (4) The constraints on tire adhesion and braking stability are shown in equation (5): (5) Where, in the formula g i The transmission ratio is... For transmission efficiency, This refers to the tire-road adhesion coefficient. This represents the vertical force corresponding to the tire. r w For tire rolling radius (for example, obtaining tire vertical force: for vehicles equipped with suspension pressure sensors, the vertical force of each wheel is calculated based on the suspension cylinder pressure and effective contact area; for vehicles without suspension pressure sensors, it is calculated based on the vehicle's static axle load and dynamic load transfer: first, the static vertical force of each axle is obtained, then the load transfer amount is calculated based on the longitudinal acceleration, and distributed to each wheel to obtain the real-time vertical force. Obtaining road surface adhesion coefficient: the vehicle controller performs online identification based on the linear relationship between the slip ratio of each drive wheel and the drive torque: when the slip ratio is within the range of 5% to 15%, multiple sets of (slip ratio, drive torque) data points are recorded, the slope of the fitted straight line is compared with the slope range corresponding to different pre-calibrated road surfaces (dry, wet, snowy, etc.) to identify the current road surface adhesion coefficient. When online identification is not possible, a preset default value is used: 0.7~0.9 for dry roads and 0.2~0.4 for wet roads).

[0027] In step two, the Model Predictive Control (MPC) optimization problem is solved offline. Based on the above model and constraints, a prediction time domain is constructed. N p The optimization problem within the system is addressed by first obtaining the vehicle speed variation curve and distance-altitude conditions over a certain future time range based on preset typical operating conditions, and then calculating the traction or braking demand sequence. Positive values ​​represent traction demand, while negative values ​​represent braking demand. Configure each power supply system SOCIETY Boundary parameters, battery power boundary, tire adhesion coefficient, torque change rate limit, and control variable relaxation are used as known inputs within the MPC prediction domain. Multiple initial simulation scenarios are constructed for different initial charge distributions and load conditions, among which the initial... SOCIETY The load distribution is set to 100%, 80%, 50%, and 30%, and the load conditions include no-load and full-load. In the k To control the cycle, we solve the problem of minimizing the following objective function, as shown in equation (6): (6) in, (7) (8) (9) (10) Equation (7) is used to balance the power sources. SOCIETY Equation (8) is used to suppress battery degradation; Equation (9) is used to smooth torque; Equation (10) is a slack variable penalty term to ensure feasibility; for SOCIETY The weighting coefficient for the deviation term can be taken as 0.4-0.6. This is the weighting coefficient for the battery degradation term, which can be set to 0.2-0.3. The weighting coefficient for the torque variation term can be 0.1-0.2, and the sum of the above weighting coefficients is 1; To achieve the above optimization objective, the decision variables are defined as follows: T(k) , slack variables are The predicted state variables are The decision variables are shown in equation (11), the slack variables are shown in equation (12), and the predicted state variables are shown in equation (13): (11) (12) (13) Among them, the running mode variable This is used to distinguish the vehicle's driving state at different times within the prediction time domain, with the drive mode corresponding to the vehicle's torque demand. Brake mode corresponds to Idle mode corresponds to a vehicle torque demand Treq≈0. This condition does not need to be included in the sample data collection and classification process, and can be controlled by the vehicle controller directly outputting the preset target torque. Operational constraints include power boundary constraints, torque boundary constraints, electrical safety boundary constraints, and tire adhesion and braking stability constraints. In an offline environment, a nonlinear optimization solver is invoked to obtain the optimal torque distribution sequence within the prediction domain, and only the first step control variable is extracted according to the rolling optimization principle. This is used for state updates, then the solution is repeated at the next time step until the complete operating cycle ends. Throughout the offline simulation, the values ​​at each time step are recorded. SOCIETY Distribution, torque demand, operating mode, optimal torque allocation ratio, and constraint activation status form a sample dataset for subsequent threshold rule extraction; Typical operating conditions can be derived from one or a combination of the following four methods: Statistical analysis of historical vehicle operation data; Modeling of standard transportation cycles with fixed operating routes; Simulation-generated working conditions based on vehicle dynamics and site parameters; Boundary scenarios are constructed for coverage, including high-load conditions, long-slope conditions, low-adhesion-coefficient road surface conditions, and single-power-source derating conditions, to ensure that the generated optimal samples are representative and complete.

[0028] Step 3 involves statistical analysis to extract mapping rules. Extract state variables at each time step from the sample dataset, including the SOC deviation of each power source. (Relative to reference value), operating mode m (drive / brake), torque requirement Constraint activation states and corresponding optimal torque distribution results Convert the optimal torque into a normalized distribution ratio. .by Based on main features, grouped by pattern, for Perform interval partitioning (binning) and define the threshold set. ,Will Divided into Equal intervals, statistical analysis within each interval r i The mean of the intervals is calculated, and amplitude limiting, normalization, and monotonicity correction are applied to obtain a stable distribution curve. For intervals with insufficient samples, they are merged with adjacent intervals to improve statistical stability. Finally, the results are solidified into a threshold rule table, which can be expressed as follows (14): (14) in, These are the allocation ratio parameters obtained through statistical extraction. An independent threshold mapping table can be constructed in braking mode. This mapping rule is an engineering expression of the optimal solution for model predictive control, used to reduce online computational complexity.

[0029] Step four involves online real-time control. When the vehicle is running, the vehicle control unit (VCU) in each control cycle k Real-time acquisition of vehicle torque demand via CAN bus Treq ( k ), each power supply SOCIETY i Current I bat,i ,Voltage V OC,i Motor speed ω i Based on information such as ΔSOC, and according to a pre-set threshold range, the value is calculated. Determine the segment interval to which the current state belongs, and then determine the interval number. index and operating mode m Then, the VCU calls the threshold rule table generated in step three and reads the allocation coefficient from the corresponding proportional lookup table structure. And calculate the target torque of each drive motor according to the following formula (15): (15) The target torque command is then output to each motor controller for execution.

[0030] By adopting this technical solution, the present invention can realize the coordinated management and control of multiple power supply systems, ensure the stable operation of vehicles under different working conditions, rationally allocate torque output, protect the safety of the power supply system, improve the rationality of energy utilization, and at the same time realize rapid online control and reduce control difficulty.

[0031] According to another embodiment of the present invention, in step one, the power boundary constraints of the power system include a maximum permissible discharge power constraint and a maximum permissible recharge power constraint, used to protect the battery from power requests exceeding its physical limits. In a practical system, the first... i Maximum permissible discharge power of the power supply system and maximum allowable recharge power The Battery Management System (BMS) calculates the maximum allowable discharge power in real time based on the current battery temperature, State of Charge (SOC), State of Health (SOH), and individual cell voltage range (for example, under normal operating conditions with room temperature and SOC of 50%~80%, the maximum allowable discharge power can be 100kW~200kW, and the maximum allowable recharge power can be 50kW~100kW; in low-temperature environments, the maximum allowable discharge power is reduced to prevent lithium plating; in the high SOC range, the maximum allowable recharge power is reduced to prevent overcharging) and sends this information to the Voltage Control Unit (VCU). For example, in low-temperature environments... Significantly reduced to prevent lithium plating; high SOC range Reduce to prevent overcharging; power boundary constraints , Maximum charging power, This represents the maximum discharge power. The torque boundary constraints for the drive motor include maximum output torque constraints and maximum braking torque constraints, used to protect the motor and inverter. The motor controller (MCU) determines the torque boundary based on the current speed. ω i Real-time calculation capability for peak torque based on DC bus voltage, temperature, etc. (Driver) and (Braking) (For example, under rated operating conditions, the maximum output torque can be selected from 2000 N·m to 4000 N·m, and the maximum braking torque can be selected from 1500 N·m to 3000 N·m. This parameter can be calculated in real time by the motor controller based on the motor speed, bus voltage, and temperature), and sent to the VCU via CAN. The power safety boundary constraints include upper and lower limits for the safe state of charge of each power system. These constraints prevent overcharging or over-discharging of the battery, extend its lifespan, and ensure safety. It is usually set at 10% to 20%. Set to 90%~95%. If the predicted SOC will reach the boundary within the predicted step size, the optimized solver automatically adjusts the torque distribution. In online control, the VCU monitors the SOC in real time. When it approaches the lower limit, the corresponding distribution ratio in the threshold rule table will be significantly reduced, or even the discharge of that branch will be temporarily stopped. The tire adhesion and braking stability constraints are determined based on the tire-road adhesion coefficient, tire vertical force, transmission ratio, and transmission efficiency (these parameters are obtained in real time according to the vehicle's operating conditions; the road adhesion coefficient can be selected from 0.2 to 0.9, the tire vertical force from 40kN to 250kN, the transmission ratio from 10 to 20, and the transmission efficiency from 0.9 to 0.98). This constraint prevents drive wheel slippage or brake wheel lock-up, ensuring dynamic stability. This constraint exists as a hard constraint in the MPC, ensuring that the torque allocated to each wheel within any prediction step does not exceed the limit provided by the current road adhesion conditions. In step four of online control, the VCU calculates the target torque... At that time, based on the current estimate and Calculate the adhesion limit torque ,like If the speed exceeds the limit, the speed will be limited to prevent slippage or lock-up.

[0032] In the above technical solution, the four constraints do not exist in isolation, but rather act together on the objective function in the MPC optimization problem and are verified sequentially in online control. For example, on a slippery road surface (low... In the uphill start-up condition, tire adhesion constraints may become active constraints, limiting the maximum driving torque; at this time, if the SOC of a certain power supply is low, the power boundary may also be constrained at the same time. The optimization solver will minimize the SOC deviation and decay penalty under the premise of satisfying these two hard constraints, and give the optimal torque distribution.

[0033] By adopting this technical solution, the present invention can clarify the specific content and determination method of each constraint, provide clear constraint standards for multi-power source collaborative control, ensure the safety of power supply and motor, and improve vehicle driving stability.

[0034] According to another embodiment of the present invention, in step two, the objective function of the model predictive control optimization problem includes a state of charge offset term (used to make each power source's SOC track the reference value to achieve power balance), a battery degradation penalty term (using the current square to approximate the degree of battery degradation and combining it with the power-torque relationship to transform it into a torque expression; this penalty term can effectively suppress high-current charging and discharging, thereby slowing down battery life degradation), a torque smoothing term (suppressing torque abrupt changes, improving driving smoothness, and reducing the impact on the transmission system), and a relaxation variable penalty term (introducing relaxation variables and setting corresponding penalty terms; the penalty term is in the form of multiplying the square of the relaxation variable by a penalty coefficient, and the penalty coefficient is greater than the weight coefficients of other terms, used to ensure that the optimization problem can still be solved when there is a constraint conflict and that the relaxation amount is within a reasonable range, ensuring that the optimization problem has a feasible solution when there is a constraint conflict). ρ For a larger penalty coefficient, generally This ensures that the optimization solution prioritizes meeting the total torque requirement, while minimizing the relaxation variable to approximate the original requirement as closely as possible. Specifically, the state-of-charge offset term minimizes the deviation between the state of charge of each power system and the preset reference value; the battery degradation penalty term approximates the degree of battery degradation using the square of the current; the torque smoothing term suppresses the rate of torque change; and the relaxation variable penalty term ensures the solvability of the optimization problem when constraints conflict. The weighting coefficients... , , The objective function is calibrated and normalized (summed to 1) to balance the importance of different optimization objectives. This objective function is then continuously optimized within the prediction time domain, achieving multi-objective collaborative optimization.

[0035] By adopting this technical solution, the present invention can achieve coordinated control of stable power, battery protection, and smooth torque by reasonably setting the components of the objective function, while ensuring the solvability of the optimization problem, improving the rationality and reliability of energy management, extending the service life of the battery and motor, and reducing equipment wear.

[0036] According to another embodiment of the present invention, the constraints of the model predictive control optimization problem further include: the relationship constraint between battery power and motor end torque, the torque change rate constraint, and the control variable relaxation constraint; the control variable relaxation constraint allows the total torque demand to be relaxed when there is a constraint conflict by introducing a relaxation variable, so as to ensure that there is a feasible solution to the optimization problem.

[0037] In the above technical solution, the relationship between battery power and motor end torque is based on the power and torque relationship models (2) and (3). The torque change rate constraint is shown in equation (15): (15) This limits the magnitude of torque variation within each control cycle, preventing mechanical or electrical shocks caused by sudden torque changes, and protecting the transmission and electrical systems. The control variable relaxation constraint is shown in equation (16): (16) (17) Introducing slack variables δ This allows for slight relaxation of the total torque requirement when necessary to ensure that a feasible solution always exists for the optimization problem, while also using a penalty term. ρ δ2 Excessive relaxation (penalty) ρ δ2 ,in ρThe penalty coefficient is greater than other weight coefficients, δ is the slack variable, and the penalty term is calculated in square form to suppress excessively large values ​​of the slack variable, making the optimization result as close as possible to the original total torque requirement. Through the cooperation of the slack variable and the penalty term, the optimization problem remains solvable under constraint conflict scenarios, while avoiding excessive relaxation of the torque requirement and maintaining the stability of vehicle power output. This mechanism ensures that when the constraints are too strong and lead to no solution, the optimizer can obtain a feasible solution by slightly sacrificing the total torque requirement, avoiding optimization failure. These constraints together constitute the feasible region of the MPC optimization problem, ensuring that the solution result satisfies both physical limits and engineering feasibility; in the model predictive control optimization problem, the relationship constraint between battery power and motor torque is added, and the discharge mode and charging mode are constrained by corresponding formulas respectively; a torque change rate constraint is added to limit the change amplitude of torque within a single control cycle to avoid rapid torque fluctuations; a control variable slack constraint is added, introducing slack variables to allow small adjustments to the total torque requirement when there is a constraint conflict, and the slack variables are set with limited ranges. By adopting this technical solution, the present invention can further improve the constraint system of model predictive control optimization problem, ensure the matching of battery power and motor torque, avoid the damage to equipment caused by sudden torque changes, and at the same time ensure the solvability of optimization problem when constraint conflicts occur, improve the stability and reliability of energy management, and ensure the safe and efficient operation of the whole vehicle system.

[0038] According to another embodiment of the present invention, step three, extracting the mapping relationship between the power state deviation and the torque distribution ratio, specifically includes: using the deviation between the power state of each power system and the preset reference value as the main feature variable, grouping the sample data according to the operating mode and constraint activation state (grouping the sample data according to the operating mode (drive / brake) and the necessary constraint activation state. For example, it can be divided into subsets such as "drive-unconstrained activation", "drive-tire adhesion constraint activation", and "brake-unconstrained activation" to capture the distribution pattern under different working conditions), dividing the power state deviation within each group into intervals, statistically analyzing the distribution characteristics of the torque distribution ratio within each interval (usually using the mean as the representative value, and combining variance information to judge stability. For intervals with insufficient samples, merging with adjacent intervals to improve statistical reliability), and processing the statistical results (limiting, normalizing, and correcting the monotonicity of the statistical results, such as as ΔSOC increases, ... r iThe distribution curve is obtained by monotonically changing (in accordance with physical laws), and this distribution curve is solidified in the form of a threshold rule table. The rule table is indexed by the power state deviation range and the operating mode for online control. Using this technical solution, the present invention can accurately extract the mapping relationship between power state deviation and torque distribution ratio. The accuracy of the mapping relationship is improved through group statistics and interval division. The solidified threshold rule table provides a fast and reliable basis for online control, reducing the computational load of online control and improving control response speed.

[0039] According to another embodiment of the present invention, in step four, after calculating the target torque of each drive motor, a real-time constraint verification and trimming step is also included: The vehicle controller applies power boundary constraints of the power system, torque boundary constraints of the drive motor, torque change rate constraints, and tire adhesion and braking stability constraints to the calculated target torque. When any constraint is triggered, the torque of each branch is limited or redistributed according to preset priority to ensure that the final output torque command meets the system safety boundary.

[0040] In the above technical solution, the power system power boundary is: based on the current battery state, the torque is inversely calculated into the power request using equation (2) or (3). and with currently allowed , In comparison, if > (Driver) or < (Braking) will either limit the amplitude proportionally or cut it directly to the boundary value; Motor torque boundary: ensure Motor peak capacity not exceeding the current speed or ; In online control, this constraint applies between the current moment and the previous moment; Tire adhesion and braking stability constraints: based on current estimates and Calculate the adhesion limit torque ,like If the limit is exceeded, the amplitude will be limited; When any constraint is triggered, the torque of each branch is redistributed or limited according to a preset priority to ensure that the final output torque command meets all system safety boundary requirements. This avoids overcurrent, overcharging, over-discharging, and drive / braking instability caused by sudden changes in operating conditions or model deviations, thereby improving system operational safety. For example, the priority can be set as: tire adhesion constraint > battery power constraint > motor torque constraint > torque change rate constraint. If multiple branches require limiting simultaneously, the remaining required torque can be redistributed proportionally.

[0041] After constraint verification and trimming, the final torque command is output and sent to the corresponding motor controller for execution via the CAN bus.

[0042] By adopting this technical solution, the present invention can perform real-time verification and adjustment of the target torque, avoiding equipment damage or vehicle instability caused by torque commands that do not meet the constraints, further improving the safety and reliability of the energy management system, and ensuring the stable operation of the whole vehicle system.

[0043] According to another embodiment of the present invention, in step four, when an abnormal state such as abnormal acquisition status, power system derating, or threshold rule table mismatch is detected, the vehicle controller switches to the equal distribution mode or the preset safe distribution mode to ensure the stable operation of the vehicle under abnormal conditions.

[0044] In the above technical solution, anomaly monitoring includes: Validity of sensor signals (e.g., whether SOC, current, and speed signals are lost or exceed limits); Power system status (such as fault codes and derating indicators reported by the BMS); Threshold rule matching status (e.g., △SOC exceeds the coverage range of the preset rule table); Abnormal Switching: When any of the following abnormal states are detected, the controller will automatically switch to the equal distribution mode or the preset safe distribution mode: sensor malfunction, power system derating (e.g., a power supply's available power drops significantly due to a fault), or threshold rule mismatch (e.g., △SOC exceeds the preset range, or the mode cannot be recognized). (When switching to the equal distribution mode or the preset safe distribution mode, a ramp function or step change rate limit is used for transition. The transition time is adaptively adjusted according to the current vehicle speed and torque change rate to avoid vehicle impact caused by sudden torque changes.) Safety mode: Even distribution mode, torque is distributed according to the proportion of available power in each power system, and if there is no available power information, it is simply distributed evenly; Preset safety distribution mode, which uses a pre-calibrated conservative distribution coefficient, such as the proportion of the rated capacity of each power source, or a fixed distribution proportion. Adaptive Updates: Under normal operating conditions, threshold parameters can be updated online or periodically calibrated and corrected based on vehicle operating data. For example, when battery degradation is detected as causing an increase in internal resistance, the degradation penalty weight or allocation curve can be adjusted appropriately, allowing the rules to adaptively evolve with battery degradation, temperature changes, and load changes, maintaining consistent control performance.

[0045] Abnormal data acquisition status includes, but is not limited to: CAN communication timeout, SOC signal switching rate exceeding a preset threshold (e.g., 10% / s), temperature sensor failure, and current sensor failure. Power system derating includes power limiting triggered by the following reasons: cell temperature exceeding a threshold (e.g., 55℃), individual cell voltage difference exceeding a threshold (e.g., 300mV), decreased insulation resistance (e.g., below 100Ω / V), and system fault diagnosis activation. Threshold rule mismatch includes: the torque obtained from the lookup table after a preset number of consecutive looksup cycles (e.g., 10 times), after real-time constraint trimming, still cannot meet the vehicle's total torque requirement, or the deviation between the actual distribution ratio after trimming and the output value of the rule table exceeds a preset threshold (e.g., 20%).

[0046] This anomaly handling mechanism ensures that the vehicle can still output executable torque commands under abnormal or faulty conditions, maintaining basic operational capabilities and improving system fault tolerance and robustness. By employing this technical solution, the present invention can promptly identify various abnormal states and prevent vehicle malfunction or equipment damage caused by abnormal states through mode switching, thereby enhancing the fault tolerance and reliability of the energy management system and ensuring stable vehicle operation under abnormal conditions.

[0047] This invention provides a multi-power source collaborative energy management system for off-highway dump trucks, comprising: Multiple independent power supply systems, each connected to a corresponding drive motor; The vehicle controller is communicatively connected to the motor controllers of the multiple power systems and each drive motor, and is used to execute the multi-power coordinated energy management method for off-highway dump trucks.

[0048] In the above technical solution, each power system consists of a battery pack, BMS, etc., and is connected to a corresponding drive motor and motor controller (MCU). The power systems are electrically isolated from each other, with no direct energy exchange. Power distribution is achieved through the vehicle controller, realizing safe redundancy and coordinated power management of multiple battery systems. It communicates with the BMS of each power system and the MCU of each drive motor via CAN bus. The VCU integrates energy management methods, collects the SOC, current, voltage of each power supply, speed of each motor, and torque demand of the whole vehicle in real time, calculates the target torque of each motor according to the threshold rule table, and sends the command to each MCU for execution after constraint verification. Each power source operates independently, while the VCU coordinates torque distribution to achieve collaborative control of the multi-power system, thus resolving issues such as dynamic deviations, reduced regenerative braking capability, and battery life degradation caused by uneven power consumption.

[0049] In the above scheme, the multi-power system consists of multiple independent power sources. Each power system operates independently without interference. Each power system is connected to a corresponding drive motor. The connection method can be direct wiring, and waterproof interfaces can be used to ensure connection stability and sealing. The power systems can be lithium battery packs, and the drive motors can be DC drive motors, both of which are readily available on the market. Regarding materials, the power system housings can be made of cold-rolled steel, the drive motor housings can be made of aluminum alloy, and the connecting wiring can be made of copper-core waterproof cables—all commonly used materials. In terms of assembly, each power system can be installed in the power compartments on both sides of the vehicle frame, and the drive motors can be installed at the vehicle's drive axle, connecting to the corresponding power system nearby to reduce wiring losses. The operating process is as follows: each power system independently provides electrical energy to its corresponding drive motor, which converts the electrical energy into mechanical energy to propel the vehicle. The power systems are not directly connected to each other, ensuring independent operation. The vehicle controller communicates with multiple power systems and the motor controllers of each drive motor. The communication method can be CAN bus communication, with a communication rate set to 500kbps to ensure fast and accurate data transmission. The vehicle controller can be an on-board ECU, the motor controller can be a DC motor controller, and the CAN bus module can be a readily available bus interface module—all commercially available equipment. In terms of materials, shielded copper core cables can be used for the CAN bus connections, and FR-4 epoxy resin boards can be used for the on-board ECU circuit boards—both commonly used materials. The vehicle controller coordinates the operation of multiple power supplies and drive motors, enabling independent power supply and coordinated torque output from multiple power supplies, avoiding direct energy flow between power sources, and improving system redundancy and stability. The operation process involves the vehicle controller receiving real-time data on the power status of each power system, the operating parameters of each drive motor, and the vehicle's torque requirements via the CAN bus. It then executes energy management methods, calculates the target torque, and outputs it to the motor controller, achieving coordinated control of multiple power supplies.

[0050] By adopting this technical solution, the present invention can build a multi-power source collaborative energy management system with a reasonable structure and reliable connection, realize the collaborative work of each power source and motor, ensure the effective implementation of the energy management method, and guarantee the stable and efficient operation of the vehicle.

[0051] According to another embodiment of the present invention, it further includes: The real-time constraint verification module is used to perform real-time constraint verification and trimming of the target torque after calculating the target torque of each drive motor. The exception handling module is used to automatically switch to the equal distribution mode or the preset safe distribution mode to distribute torque when an abnormal state is detected.

[0052] In the above technical solution, the real-time constraint verification module is integrated into the VCU and is used to perform real-time verification and trimming of the power boundary, torque boundary, torque change rate and attachment constraint of the torque command after calculating the target torque. The anomaly handling module, integrated into the VCU, automatically switches to either the equal-sharing mode or a preset safe-sharing mode when abnormal conditions such as sensor malfunction, power derating, or rule mismatch are detected. This module enables anomaly monitoring, mode switching, and adaptive updates, ensuring the system can still output executable torque commands under fault conditions, thus enhancing system fault tolerance and robustness.

[0053] These two modules work together to ensure that the system can stably and reliably achieve multi-power source collaborative energy management under both normal and abnormal conditions, thus fully realizing the technology chain from offline optimization to online control.

[0054] In the above scheme, the real-time constraint verification module is used to perform real-time constraint verification and trimming of the target torque after calculating the target torque of each drive motor. This module can be integrated into the vehicle controller, sharing a circuit board with the vehicle controller, without occupying additional installation space. In terms of equipment selection, the real-time constraint verification module can use readily available logic control chips, all of which are commercially available devices. The specific working process of the real-time constraint verification module is as follows: it receives the target torque data transmitted by the vehicle controller, and at the same time calls the real-time parameters transmitted by the power sensor, torque sensor, road condition sensor, etc., compares the parameters such as power, torque change rate, and tire adhesion corresponding to the target torque with preset constraint thresholds. When any parameter is detected to exceed the constraint threshold, it immediately sends a constraint trigger signal to the vehicle controller and proposes torque adjustment suggestions according to preset priorities, cooperating with the vehicle controller to complete torque trimming or redistribution. After adjustment, it verifies again until the target torque meets all constraint requirements. The anomaly handling module automatically switches to either an equal-sharing mode or a preset safe-distribution mode for torque allocation when an abnormal state is detected. This module can also be integrated into the vehicle controller, working in conjunction with the real-time constraint verification module and sharing the data acquisition and storage resources of the vehicle ECU. The real-time constraint verification module and the anomaly handling module can share the data acquisition and storage resources of the vehicle controller, collaboratively completing constraint protection and fault tolerance processing. This improves the system's protection and fault tolerance functions, enabling the system to stably complete torque distribution under both normal and abnormal operating conditions, reducing the risk of equipment damage. For equipment selection, the anomaly handling module can use readily available microprocessors, all of which are commercially available devices. The operation process involves the anomaly handling module monitoring the sensor data, power operating parameters, and threshold rule matching status transmitted by the vehicle ECU in real time. When it detects anomalies in data acquisition, power derating, or rule mismatch, it immediately triggers a mode switching command, controlling the vehicle controller to switch to either an equal-sharing mode or a preset safe-distribution mode to ensure stable vehicle operation. By adopting this technical solution, the present invention can further improve the function of the energy management system by adding two functional modules, enhance the system's constraint verification capability and anomaly handling capability, ensure the system can operate stably under various working conditions, reduce the risk of equipment damage, and ensure vehicle driving safety.

[0055] This invention also provides an off-highway dump truck, including the aforementioned off-highway dump truck multi-power source collaborative energy management system. The various components of the system are assembled and connected according to the vehicle structure and installation requirements. This enables the off-highway dump truck to operate stably in off-highway mining scenarios, achieving balanced power supply and reasonable torque distribution from multiple power sources, adapting to various working conditions such as heavy loads, climbing, and braking, and extending the service life of the power supply and transmission components.

[0056] Although embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. They can be applied to various fields suitable for the present invention. For those skilled in the art, other modifications can be easily made. Therefore, without departing from the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details and embodiments shown and described herein.

Claims

1. A multi-power source collaborative energy management method for off-highway dump trucks, characterized in that, include: Step 1: Construct a multi-power system model for model predictive control and set the operating constraints of the multi-power system. The multi-power system model includes the power state model of each power system, the power-torque relationship model between each power system and the corresponding drive motor, and the vehicle traction and braking demand model. The vehicle traction and braking demand model is used to characterize the torque demand changes under the preset route. The operating constraints include power boundary constraints of the power system, torque boundary constraints of the drive motor, electrical safety boundary constraints, and tire adhesion and braking stability constraints. Step 2: Based on the multi-power system model and operating constraints, construct the model predictive control optimization problem of the multi-power system in the prediction time domain; in offline mode, perform rolling optimization to solve the optimization problem based on preset typical operating conditions, obtain the optimal torque distribution sequence in the prediction time domain, and record the torque distribution value, power status, operating mode and constraint activation status at each time in the optimal torque distribution sequence to form a sample dataset. Step 3: Perform statistical analysis on the sample dataset, extract the mapping relationship between the power status value of each power system and the preset reference value and the torque distribution ratio, and solidify the mapping relationship into a threshold rule table indexed by the power status deviation range and the operating mode. Step 4: When the vehicle is running, the vehicle controller collects the vehicle torque demand, the current power status of each power system and the current operating mode in real time. It calculates the power status deviation of each power system based on the current power status, and queries the threshold rule table based on the power status deviation and the current operating mode to obtain the target torque distribution ratio of each power system. It calculates the target torque of each drive motor based on the target torque distribution ratio and the vehicle torque demand, and outputs the target torque command to each motor controller for execution.

2. The multi-power source collaborative energy management method for off-highway dump trucks as described in claim 1, characterized in that, In step one, the power boundary constraints of the power system include the maximum allowable discharge power constraint and the maximum allowable recharge power constraint; The torque boundary constraints of the drive motor include the maximum output torque constraint and the maximum braking torque constraint; The electrical safety boundary constraints include the upper and lower limits of electrical state safety constraints for each power system. The constraints on tire adhesion and braking stability are determined based on the tire-road adhesion coefficient, tire vertical force, transmission ratio, and transmission efficiency.

3. The multi-power source collaborative energy management method for off-highway dump trucks as described in claim 2, characterized in that, In step two, the objective function of the model predictive control optimization problem includes a state of charge offset term, a battery degradation penalty term, a torque smoothing term, and a slack variable penalty term. Among them, the state of charge offset term is used to minimize the deviation between the state of charge of each power system and the preset reference value, the battery degradation penalty term approximates the degree of battery degradation with the square of the current, the torque smoothing term is used to suppress the torque change rate, and the slack variable penalty term is used to ensure the solvability of the optimization problem when there are constraint conflicts.

4. The multi-power source collaborative energy management method for off-highway dump trucks as described in claim 3, characterized in that, The constraints of the model predictive control optimization problem also include: the relationship between battery power and motor torque, the torque change rate constraint, and control variable relaxation constraints; the control variable relaxation constraints allow for relaxation of the total torque demand when constraints conflict, so as to ensure that there is a feasible solution to the optimization problem.

5. The multi-power source collaborative energy management method for off-highway dump trucks as described in claim 1, characterized in that, Step 3, extracting the mapping relationship between the power status deviation and the torque distribution ratio, specifically includes: taking the deviation between the power status of each power system and the preset reference value as the main feature variable, grouping the sample data according to the operating mode and constraint activation state, dividing the power status deviation within each group into intervals, statistically analyzing the distribution characteristics of the torque distribution ratio within each interval, processing the statistical results to obtain the distribution curve, and solidifying the distribution curve in the form of a threshold rule table for online control.

6. The multi-power source collaborative energy management method for off-highway dump trucks as described in claim 4, characterized in that, Step four, after calculating the target torque of each drive motor, also includes a real-time constraint verification and trimming step: The vehicle controller applies power boundary constraints of the power system, torque boundary constraints of the drive motor, torque change rate constraints, and tire adhesion and braking stability constraints to the calculated target torque. When any constraint is triggered, the torque of each branch is limited or redistributed according to preset priority to ensure that the final output torque command meets the system safety boundary.

7. The multi-power source collaborative energy management method for off-highway dump trucks as described in claim 1, characterized in that, In step four, when abnormal conditions such as abnormal data acquisition status, power system derating, or threshold rule mismatch are detected, the vehicle controller switches to the equal distribution mode or the preset safe distribution mode to ensure the stable operation of the vehicle under abnormal conditions.

8. A multi-power cooperative energy management system for off-highway dump trucks used to perform the method as described in any one of claims 1-7, characterized in that, include: Multiple independent power supply systems, each connected to a corresponding drive motor; The vehicle controller is communicatively connected to the motor controllers of the plurality of power systems and each drive motor, and is used to execute the off-highway dump truck multi-power collaborative energy management method as described in any one of claims 1 to 7.

9. The off-highway dump truck multi-power source collaborative energy management system as described in claim 8, characterized in that, Also includes: The real-time constraint verification module is used to perform real-time constraint verification and trimming of the target torque after calculating the target torque of each drive motor. The exception handling module is used to automatically switch to the equal distribution mode or the preset safe distribution mode to distribute torque when an abnormal state is detected.

10. An off-highway dump truck, comprising the off-highway dump truck multi-power source collaborative energy management system as described in claim 9.

Images