Digital-twin-based power real-time control system and method for extended-range tracked tractor

Abstract

The application relates to the field of tractor power control, in particular to a range-extending crawler tractor power real-time control system and method based on digital twinning. The system comprises a physical power unit, a data interaction unit, a digital twinning unit and a decision control unit. The physical power unit collects operation state parameters and working condition data, which are sent to the digital twinning unit after time sequence alignment by the data interaction unit. The digital twinning unit models and deduces the working condition and generates target and constraint parameters, and then the decision control unit generates control instructions. Finally, the physical power unit regulates and controls power output and energy distribution. The application relies on digital twinning virtual-real synchronization and prospective working condition deduction, successfully realizes the scheme purpose of precise regulation and control of hilly and mountainous electric crawler tractor power, effectively solves the problems of traditional control lag and poor adaptability, prolongs the service life of power components, improves the system operation robustness and energy utilization efficiency, and highly adapts to complex field operation scenes with local terrain undulation and load mutation.

Description

Technical Field

[0001] This application relates to the field of tractor power control, and in particular to a real-time power control system and method for a range-extended tracked tractor based on digital twins. Background Technology

[0002] High-intensity field operations such as plowing, rotary tilling, ditching, and harvesting in hilly and mountainous areas heavily rely on tracked tractors. Traditional diesel direct-drive models suffer from high fuel consumption, high emissions, and significant noise. Pure electric tractors can achieve zero emissions, but their range and output power are limited by battery performance and field charging conditions, making it difficult to meet the demands of large-scale continuous operations. Range-extended tractors balance range and power output, making them more suitable for the core needs of large-scale, high-intensity continuous operations in hilly and mountainous areas. However, the complex terrain of hilly and mountainous areas, high track resistance, and high steering energy consumption mean that the tractor's power demand varies strongly non-linearly with working conditions. In addition, the complex multi-power source coordination control of range-extended tractors makes it difficult to accurately match power supply with real-time working conditions, resulting in the energy efficiency and operational performance of the range-extended system being difficult to fully realize.

[0003] In response, some existing technologies improve the matching accuracy between power supply and real-time operating conditions through multi-energy dynamic collaborative control, thereby releasing the energy efficiency and operational performance of range-extended systems. For example, the multi-energy collaborative system for range-extended tracked tractors disclosed in Chinese patent application CN121626090A uses a vehicle controller to build a collaborative architecture, adopts a dynamic programming energy management strategy, combines the Pontryagin minimum principle to solve for optimal power allocation and dynamically adjust weight coefficients; it estimates the road adhesion coefficient through Kalman filtering to achieve adaptive torque distribution between the left and right tracks; and it is supplemented by fuzzy adaptive PID generator control, sliding mode variable structure clutch control, and fault prediction models to collaboratively schedule the engine, generator, electric drive, hydraulic, and PTO subsystems, optimizing power and torque output, improving energy efficiency, traction stability, and operational adaptability under complex operating conditions, deeply exploring the multi-source power collaborative potential of range-extended systems, and better leveraging their energy efficiency and operational performance.

[0004] However, this scheme employs a feedback-based energy management strategy based on real-time sensor data, which suffers from significant control lag. Its control logic, relying on the Pontryagin minimum principle, requires dynamic adjustments to weighting coefficients and power distribution schemes only after vehicle tilt angles, speed fluctuations, or sudden load changes occur. While this approach can achieve local optimization of power distribution and energy consumption control under real-time operating conditions, enabling the vehicle's power system to quickly respond to changes in operating conditions, in complex scenarios such as frequent changes in slope gradients in hilly terrain and sudden load variations, lag adjustment easily leads to delayed power response and inaccurate energy matching. This, in turn, causes problems such as increased energy consumption, track slippage, and decreased operational smoothness, significantly reducing the adaptability of the control strategy to different operating conditions. Summary of the Invention

[0005] This application provides a real-time power control system and method for range-extended tracked tractors based on digital twins, in order to solve the problem that the control lag caused by the feedback optimization method that relies on real-time sensor data leads to low adaptability of the power control strategy for range-extended electric tracked tractors under working conditions.

[0006] To solve the above-mentioned technical problems, this application provides the following technical solution: The real-time power control system for range-extended tracked tractors based on digital twins includes: The physical power unit includes various actuators mounted on the tractor body, each actuator including a range extender, a battery system, a PTO motor, a power conversion device, and a track drive motor; it is used to collect the operating status parameters of each actuator in real time; it also includes a working condition acquisition device for collecting working condition data in real time, the working condition data including terrain data and work load data; The data interaction unit is used to acquire and time-series align operating condition data and running status parameters; The digital twin unit constructs a virtual power model based on the hardware structure, parameter characteristics, and operating rules of each actuator, and constructs a working condition digital model based on the temporal relationship of terrain data and work load data. It acquires real-time operating status parameters and real-time working condition data according to the acquisition time, and corrects the virtual power model based on the real-time operating status parameters to obtain a fitted power model. It inputs the real-time working condition data into the working condition digital model and obtains the working condition data within a preset prediction window as predicted working condition data. Based on the real-time working condition data and predicted working condition data, it performs simulation analysis by fitting the power model and the working condition digital model to obtain the tractor's power demand, energy flow between actuators, and power distribution. The obtained data is then processed into target parameters and constraint parameters, which are sent to the data interaction unit. The decision control unit receives the target parameters and constraint parameters sent by the data interaction unit, and performs fusion processing with real-time operating data and preset safety constraints to obtain control commands; the control commands are then sent to the physical power unit through the data interaction unit. After receiving the control command, the physical power unit controls the power output and energy distribution of each actuator according to the control command.

[0007] The basic principle and beneficial effects of this solution are as follows: This application collects the operating status parameters of actuators such as range extenders and battery systems, as well as terrain and workload-related operating condition data, through a physical power unit. The timing alignment of operating condition data and operating status parameters is completed through a data interaction unit. A virtual power model matching hardware characteristics and a timing-related digital operating condition model are constructed based on a digital twin unit. A fitted power model is obtained by correcting the model through real-time operating parameters. Then, based on real-time operating conditions, the operating conditions are predicted and simulation analysis is carried out to form target parameters and constraint parameters. Control commands are generated by the decision control unit in combination with safety constraints. Finally, the physical power unit regulates the power output and energy distribution of each actuator according to the commands. In the operating scenario of hilly and mountainous terrain with frequent undulations and sudden changes in operating load, the principle of virtual-real synchronous modeling and forward-looking operating condition simulation is used to realize the advance prediction of power demand, solve the problem of power distribution imbalance caused by control lag, and ensure that power output and energy distribution always match the actual operating needs, thereby improving the control accuracy and operational stability under complex operating conditions.

[0008] This application reduces the instantaneous load impact and frequent drastic adjustments of core power components by predicting power demand in advance and optimizing power allocation. This effectively extends the stable service life of the power battery and various actuators, while avoiding heat accumulation caused by abnormal current fluctuations, reducing the risk of operational anomalies caused by component overheating, and optimizing the overall operating thermal state. Compared with feedback energy management strategies based on real-time sensor data, this application can correct the power allocation method online according to the performance degradation and parameter drift of each actuator. In long-term, high-intensity, continuous operation scenarios such as plowing and rotary tilling in hilly areas, it can continuously ensure that the power components maintain a stable state during long-term operation, fundamentally alleviating the problem of decreased control adaptability caused by component aging.

[0009] This application integrates control commands with a decision control unit and safety constraints to ensure that power regulation remains within a safe and stable range. Even in the event of sudden and drastic changes in operating conditions, it guarantees a smooth transition in power output and energy distribution, significantly improving the fault tolerance and operational robustness of the regulation process. Regulation methods that rely solely on real-time operating condition signals for passive response are ill-suited to anticipating and responding to multiple disturbances such as frequent undulations in hilly terrain and sudden changes in field workloads. This can easily lead to power regulation lag or even instability. This application can readily handle various sudden changes in operating conditions and maintain the stable operation of the power system at all times.

[0010] This application, by collecting real-time operating condition information and dynamically predicting changing trends, ensures that the power control strategy continuously adapts to the operational needs of different terrains and load conditions. Simultaneously, it optimizes energy flow and power distribution through simulation analysis, eliminating ineffective energy loss, improving operational efficiency, and reducing overall energy consumption. Static control methods using offline simulation calibration rely on simulation models only for offline verification during the product design phase. These models cannot be integrated with real-world operating data and dynamically optimized control strategies in complex hilly and mountainous terrains with varying soil resistance and slopes. This approach is prone to problems such as delayed power response, unreasonable power distribution, and low energy utilization efficiency. This application, through forward-looking control, completely solves these shortcomings, comprehensively improving operational quality and energy utilization levels under complex conditions.

[0011] In summary, this application, relying on the synchronization of digital twins with real and virtual systems and forward-looking working condition simulation, successfully achieves the goal of precise power control for electric tracked tractors in hilly and mountainous areas. It effectively solves the problems of traditional control lag and poor adaptability, extends the service life of power components, improves the robustness of system operation and energy utilization efficiency, and is highly adaptable to complex field operation scenarios with undulating terrain and sudden load changes.

[0012] Furthermore, the digital twin unit is also used to construct a range extender generator model, a power battery model, a drive motor model, and a power distribution model based on the hardware structure of each actuator. The range extender generator model is used to characterize the changes in the generator's power output and efficiency under different operating conditions. The power battery model is used to characterize the changes in the battery's state of charge, charging and discharging power limitations, and dynamic response characteristics. The drive motor model is used to characterize the torque output characteristics and efficiency distribution of the tracked drive motor. The power distribution model is used to characterize the power coordination relationship and system constraints between the power sources of each actuator. The range extender generator model, power battery model, drive motor model, and power distribution model are integrated to form a virtual power model.

[0013] This application constructs a virtual power model by splitting and constructing a range extender generator model, a power battery model, a drive motor model, and a power distribution model, and then merging them. This allows for the precise decomposition of the independent operating characteristics of each actuator, achieving a refined representation of the operating conditions of individual components. Compared to the overall modeling approach, the layered modeling and fusion mode reduces parameter coupling interference and facilitates the individual calibration of the operating parameters of each power source. In hilly, heavy-load, uphill operation scenarios, it can independently capture real-time changes in generator power generation efficiency and motor torque output. Relying on the interconnected characteristics of each sub-model, it can pre-match the power coordination relationship of multiple power sources, avoiding power output jerks caused by imbalances in a single parameter, and improving the accuracy and smoothness of power matching in complex terrain.

[0014] Furthermore, the terrain data includes slope information, surface adhesion index, and equivalent resistance coefficient; the slope information constrains the value range of the surface adhesion index; the surface adhesion index is combined with the preset inherent characteristics of field soil to correlate the variation law of the equivalent resistance coefficient; the terrain data of each region are coupled and correlated in a time sequence; the working condition digital model, based on the coupled terrain data and the working load data, quantifies the relationship between slope change, surface adhesion conditions, soil resistance, and working load on the overall power demand of the machine during hilly and mountainous operations; the working condition digital model uses slope information, equivalent resistance coefficient, and adhesion index as input parameters to analyze and obtain the equivalent resistance characterization and load demand characterization under the corresponding working scenario, and obtains the fitted power model through the virtual power model corrected by the equivalent resistance characterization and load demand characterization.

[0015] This application constrains and temporally couples the values ​​of slope information, surface adhesion index, and equivalent resistance coefficient. It quantifies the impact of multiple terrain elements on power demand using a digital working condition model, accurately extrapolating and correcting the equivalent resistance and load demand representations to obtain a fitted dynamic model. The temporal coupling of multiple terrain parameters eliminates the lag bias of single-point terrain data, adapting to operational scenarios with uneven terrain and varying soil textures in farmland. It can dynamically correct model accuracy in real-time to adapt to complex field topography changes, eliminating the need for manual calibration of terrain parameters. Furthermore, it can predict changes in terrain resistance based on parameter correlation patterns, providing preliminary data support for subsequent power allocation and enhancing the adaptive capability for working condition adaptation.

[0016] Furthermore, the digital twin unit includes state variables to characterize the operating status and parameter characteristics of the virtual power model and the working condition digital model. The physical power unit extracts key state parameters from the operating status parameters and working condition data, and transmits these key state parameters to the digital twin unit in real time via the tractor's CAN bus. Upon receiving the key state parameters, the digital twin unit uses them to correct the state variables of the virtual power model and the working condition digital model. The virtual power model derives predicted operating results corresponding to power demand, energy flow, and power distribution based on the corrected state variables. The physical power unit collects the actual operating results of each actuator of the tractor under actual working conditions and sends them to the digital twin unit. The digital twin unit compares the predicted operating results derived from the virtual power model with the actual operating results collected by the physical power unit to obtain the operating deviation. Based on the hardware structure, parameter characteristics, and operating rules of each actuator on which the virtual power model is based, the digital twin unit extracts the model structure parameters, characteristic fitting parameters, and working condition correlation parameters during the process of fitting the virtual power model to form the fitted power model, and uses these as internal key parameters. The digital twin unit performs online calibration of the internal key parameters built into the virtual power model based on the operating deviation.

[0017] This application sets up dedicated state variables to represent the model's operational characteristics, transmits key state parameters via a CAN bus to update the model's state, and compares the predicted results with actual operational deviations to calibrate internal key parameters online, allowing the digital model of the working conditions to continuously evolve with the equipment and environment. Based on the principle of deviation closed-loop calibration, it can automatically compensate for model inaccuracies caused by aging of mechanical parts and drift in field working conditions, adapting to long-term uninterrupted farmland operation scenarios; it not only ensures a high degree of consistency between virtual simulation results and actual working conditions, but also eliminates the need for offline recalibration of model parameters, continuously maintaining simulation accuracy, while simultaneously achieving synchronous iteration of state variables and internal key parameters, improving the model's long-term adaptability.

[0018] Furthermore, the digital twin unit constructs a virtual power model and a working condition digital model using a reduced-order model or a proxy model; the digital twin model sets its calculation cycle according to the timing rhythm of the tractor's overall control cycle, and the calculation cycle is slower than the timing rhythm of the tractor's overall control cycle; within the calculation cycle, the digital twin unit synchronously updates the state variables and internal key parameters of the virtual power model and the working condition digital model, and generates optimization decision-related parameters including target parameters and constraint parameters based on the updated state variables and calibrated internal key parameters, and sends the optimization decision-related parameters to the decision control unit.

[0019] This application employs a reduced-order model or a surrogate model to build a twin model, setting a computation cycle slower than the vehicle control cycle, synchronously updating model parameters, and generating and distributing optimization decision-related parameters. The reduced-order and surrogate models can simplify high-order computational dimensions and reduce computational power consumption, adapting to application scenarios where the computing power of on-board computing units at the field edge is limited. The staggered cycle operation mode avoids timing conflicts in real-time vehicle control, neither interfering with the underlying power real-time regulation nor reserving computation time for upper-level optimization decisions, stably outputting target and constraint parameters, and balancing computational real-time performance, computational power adaptability, and decision reliability.

[0020] Furthermore, the decision control unit performs boundary constraint checks, rate of change constraint judgments, and safety condition verifications on the target parameters and constraint parameters according to preset verification conditions. When the target parameters and constraint parameters meet the preset verification conditions, the decision control unit sets the target parameters and constraint parameters as the vehicle target control parameters and generates control commands based on the vehicle target control parameters. When the target parameters and constraint parameters do not meet the preset verification conditions or when complete target parameters and constraint parameters are not received within the preset timeout window, the preset local control strategy is invoked to generate backup control parameters, and control commands are generated based on the backup control parameters.

[0021] This application performs boundary, rate of change, and safety checks on target and constraint parameters. Compliant parameters directly participate in command generation, while abnormal parameters or communication timeouts trigger a switch to a local strategy to generate backup parameters. In scenarios where communication links are unstable due to tree cover or electromagnetic interference, the control risks caused by twin parameter failures can be automatically avoided. The multi-check mechanism can intercept parameter inputs that exceed the range or exhibit abrupt changes, preventing power control from exceeding limits. The dual-mode parameter switching logic requires no manual intervention, ensuring the continuity and safety redundancy of power control in complex electromagnetic environments.

[0022] Furthermore, the digital twin unit extracts the variation characteristics and periodic characteristics of historically collected operating status parameters and operating condition data as verification characteristics, and extracts the variation characteristics and periodic characteristics of the operating condition data and operating status parameters received in the current calculation cycle as comparison characteristics. When the difference between the comparison characteristics and the verification characteristics is greater than the preset verification difference, the digital twin unit combines the historically collected operating status parameters and operating condition data with the virtual power model and the operating condition digital model to verify and compensate the state variables and internal key parameters of the virtual power model and the operating condition digital model in the current calculation cycle. The compensated state variables and internal key parameters are used as substitute parameters and sent to the decision control unit. The decision control unit generates control commands based on the received substitute parameters.

[0023] This application extracts historical and current operating conditions, changes in operating parameters, and cyclical characteristics for difference comparison. When discrepancies exceed limits, it combines a twin model to compensate for state variables and internal key parameters and outputs alternative parameters. Based on the principle of identifying data anomalies through time-series feature comparison, it can accurately identify instantaneous data distortion and minor sensor drift caused by bumpy field operations, adapting to the undulating operating scenarios of muddy fields. Data self-healing compensation can be completed without hardware maintenance, and alternative parameters can be seamlessly integrated with control logic, avoiding power regulation fluctuations caused by instantaneous data anomalies and improving operational stability under harsh conditions.

[0024] Furthermore, the physical power unit also includes temperature sensors that monitor the battery system temperature and ambient temperature. The physical power unit sends the battery system temperature and ambient temperature to the digital twin unit through the data interaction unit. The digital twin unit uses the total energy consumed by each actuator during the tractor's operation as the overall energy consumption and the remaining energy that the battery system can currently output as the battery energy reserve. Based on the received battery system temperature and ambient temperature, the digital twin unit simulates the correlation between the overall energy conversion process and the battery system temperature under different operating environment temperatures. It limits the battery system energy conversion power according to a preset battery safety temperature threshold and predicts the overall operating efficiency through real-time operating environment temperature. The digital twin unit constructs a dynamic battery temperature model using operating environment temperature, overall operating efficiency, energy conversion time, and battery system temperature changes caused by efficiency changes as parameters, and manages battery operating parameters by setting battery temperature constraints. The digital twin unit combines the dynamic battery temperature model to predict energy conversion conditions and battery system temperature change trends, performs logical calculations based on the tractor's real-time power requirements, formulates an adaptive energy conversion strategy based on the calculation results, and controls the timing of battery system discharge and the charging time of the battery system through the energy conversion strategy.

[0025] This application adds a temperature sensor to collect battery and ambient temperatures, defines the overall energy consumption and battery energy storage, builds a dynamic battery temperature model, and formulates an adaptive charge-discharge conversion strategy. It can adapt to cross-seasonal operation scenarios with significant temperature differences between winter and summer in mountainous areas. Based on the correlation mechanism between temperature and energy conversion, it limits the battery conversion power with temperature constraints, taking into account both operational safety and battery life. It can also calculate the charging and discharging timing based on power demand, automatically matching the energy scheduling logic for different working conditions such as field operations, idling, and uphill loads, to achieve intelligent energy distribution under temperature linkage.

[0026] Furthermore, the digital twin unit fits the overall energy consumption variation pattern corresponding to different operating environment temperatures, and combines the real-time operating environment temperature and real-time battery system temperature to dynamically correct the battery energy storage under different operating conditions based on the overall energy consumption variation pattern.

[0027] This application fits the energy consumption patterns of the entire machine under different ambient temperatures and dynamically corrects the battery energy inventory under multiple operating conditions by combining ambient and battery temperatures. It adapts to all-day crop rotation operation scenarios with varying day-night temperature gradients in farmland, eliminating the bias in battery remaining energy assessment caused by temperature fluctuations. Based on the principle of energy consumption temperature correlation fitting, it accurately corrects the inventory estimation errors caused by low-temperature capacity decay and high-temperature energy consumption increases. It does not require fixed calibration coefficients and adaptively adapts to all-weather temperature changes, providing accurate inventory data support for energy scheduling in long-term operations.

[0028] This application fits the energy consumption variation patterns of the entire machine under different operating ambient temperatures, and combines real-time operating ambient temperature and real-time battery system temperature to dynamically correct battery energy reserves under different operating conditions based on the energy consumption variation patterns of the entire machine. By utilizing the inherent correlation mechanism between ambient temperature, battery temperature, and overall energy consumption, it can eliminate the inventory estimation deviation caused by battery capacity decay and energy consumption fluctuations under high and low temperature conditions, adapting to the actual scenarios of gradual temperature changes in the field at different times and continuous operation in all weather conditions; it can overcome the limitations of fixed calibration parameters, adaptively correct the energy reserve calculation distortion caused by temperature disturbances, and provide accurate and reliable inventory data for the overall machine energy conversion strategy and charge / discharge timing control, improving the rationality of energy scheduling and the accuracy of range prediction under complex temperature conditions. Attached Figure Description

[0029] Figure 1 A schematic diagram of the activities of each unit in the real-time power control system of a range-extended tracked tractor based on digital twin; Figure 2 A schematic diagram of the control architecture of a real-time power control system for a range-extended tracked tractor based on digital twins; Figure 3 A flowchart illustrating the power control method for range-extended tracked tractors based on digital twins; Figure 4 This is a flowchart illustrating the process of controlling the entire vehicle based on operating status parameters during operation. Detailed Implementation

[0030] The following will describe the concept and technical effects of this application clearly and completely with reference to embodiments, so as to fully understand the purpose, features and effects of this application. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. Other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are all within the scope of protection of this application. Example 1

[0031] In this embodiment, a digital twin platform is provided, deployed on a cloud server, which is communicatively connected to the range-extended electric tracked tractor. During actual tractor operation, the operating status parameters of the physical power unit (including the engine speed, generator output power, start-stop status, and operating point efficiency of the range extender; the state of charge, temperature, and charging / discharging power of the power battery; and the speed, torque, and power of the drive motor and PTO work motor) and environmental parameters (including terrain slope, workload, soil resistance, and ambient temperature) are collected in real time via the tractor's onboard CAN bus and transmitted to the digital twin platform. This data is used to drive the virtual power model to update its virtual state (the virtual state includes the virtual power model, the operating condition digital model, and the built-in range extender dynamic model. Based on the real-time operating status parameters, operating condition parameters, battery temperature, and ambient temperature uploaded from the tractor's physical terminal, the platform updates the real-time operating characteristics of each model in real time. The range extender dynamic model can synchronously update the virtual operating point, economic operating range, and power generation efficiency characteristics of the range extender). The digital twin platform, based on the updated virtual power model and combined with the operating characteristics of the range extender (such as efficiency MAP),... The system uses curves, power generation response characteristics, and start-stop switching logic to predict and calculate power demand, energy flow, and power distribution in the short term. It dynamically generates target operating parameters (including the target operating speed of the range extender, target power generation, upper limit of charging and discharging power of the power battery, and power distribution ratio of the drive motor and PTO working motor) and constraint parameters (including the upper limit of total system power, dynamic safety margin, and energy management constraints) for vehicle control.

[0032] The digital twin platform does not directly participate in the underlying execution control, but rather acts as the optimization decision-making layer for tractor power, outputting optimized control parameters to the vehicle controller. Under the premise of meeting real-time performance, safety, and system constraints, the vehicle controller determines the effectiveness of the optimized control parameters, integrates and processes them, and executes them. This achieves dynamic matching and real-time optimized operation of the tractor power in complex hilly conditions without altering the original control architecture of the vehicle. By constructing a virtual-physical closed-loop collaborative mechanism between physical power, operating environment, digital twin model, and vehicle controller, the system achieves forward-looking adjustment and continuous optimization of the power control system for range-extended electric tracked tractors in hilly terrain, improving the tractor power control system's adaptability and operational stability to changes in slope, load fluctuations, and operating environment.

[0033] like Figure 1 As shown, the real-time power control system for range-extended tracked tractors based on digital twins includes a physical power unit, a data interaction unit, a digital twin unit, and a decision control unit.

[0034] The physical power unit, located on the chassis of the hilly and mountainous range-extended electric tracked tractor, realizes the vehicle's walking, working, and energy conversion functions. The physical power unit includes various actuators mounted on the tractor chassis, as well as the execution and energy units for each actuator. The actuators include a range extender, battery system, PTO motor, power conversion device, and track drive motor. The range extender is an independent onboard power generation unit that does not directly drive the tracks and working mechanisms; it only generates power through fuel to replenish the battery system or directly supply power to the motor, extending the vehicle's range in a hybrid mode. It is used to collect real-time operating status parameters of each actuator. It also includes a working condition data acquisition device for real-time collection of working condition data, including terrain data and workload data. Under the control of the vehicle controller, the physical power unit executes specific power output and energy distribution operations. Its operating status parameters serve as the input data source for the digital twin unit and also as feedback on the control execution effect.

[0035] The data interaction unit is used to acquire and time-align operating condition data and running status parameters. It enables bidirectional data communication between the physical powertrain system, vehicle controller, and digital twin platform. Based on the vehicle's CAN bus, the data interaction unit facilitates the exchange of various data, supporting both periodic and event-triggered transmissions to ensure the synchronization of the virtual powertrain model with the running status parameters of the physical powertrain unit.

[0036] The digital twin unit, deployed within the digital twin platform, constructs a virtual power model based on the hardware structure, parameter characteristics, and operational patterns of each actuator. It also constructs a working condition digital model based on the temporal relationship between terrain data (including slope information, surface adhesion indicators, and equivalent drag coefficients) and operational load data. The digital twin unit acquires real-time operating status parameters and real-time working condition data based on the acquisition time. It then corrects the virtual power model using these parameters to obtain a fitted power model. Real-time working condition data is input into the working condition digital model, and the unit obtains the working condition data within a preset prediction window (set by the administrator based on the tractor's terrain change rate in hilly and mountainous areas, overall load fluctuation cycle, motor and battery dynamic response delay, and vehicle control calculation cycle) as predicted working condition data. Based on the real-time and predicted working condition data, simulation analysis is performed using the fitted power model and the working condition digital model to obtain the tractor's power demand, energy flow and power distribution among the actuators, and simultaneously matches the coordination between the range extender's power generation and battery power supply. The obtained data is processed into target parameters and constraint parameters, which are then sent to the data interaction unit.

[0037] The digital twin unit constructs a range extender generator model, a power battery model, a drive motor model, and a power distribution model based on the hardware structure of each actuator. The range extender generator model is used to characterize the changes in the generator's power output and efficiency under different operating conditions. The power battery model is used to characterize the changes in the battery's state of charge, charging and discharging power limitations, and dynamic response characteristics. The drive motor model is used to characterize the torque output characteristics and efficiency distribution of the track drive motor. The power distribution model is used to characterize the power coordination relationship between the power sources of each actuator and the system constraints. The range extender generator model, power battery model, drive motor model, and power distribution model are integrated to form a virtual power model.

[0038] The decision control unit, implemented by the tractor's vehicle controller, receives target and constraint parameters from the data interaction unit. It then combines these parameters with real-time operating data and preset safety constraints (set by the administrator based on the rated operating thresholds, temperature rise, charge / discharge rates, and overload protection limits of the range extender, battery system, drive motor, and power conversion device; combined with the operating limits for hilly and mountainous terrain with heavy loads and bumpy, variable load conditions; and the vehicle's electrical safety specifications and mechanical power output limits; these constraints define the operating boundaries and power adjustment range of each component) to determine their effectiveness and generate control commands. These commands are then sent to the physical power unit via the data interaction unit. Under the premise of ensuring control stability, real-time performance, and safety, the decision control unit converts the target and constraint parameters into specific control commands and applies them to the actuators in the physical power unit. When the digital twin platform outputs an abnormality or communication is interrupted, the decision control unit automatically switches to a preset local control strategy to ensure the continuity and safety of the vehicle's operation.

[0039] The validity assessment includes verifying the numerical range, rate of change, message integrity, and timing synchronization of target parameters, constraint parameters, and operating condition data; checking the rationality of parameter logical relationships; and eliminating invalid data that is out of range, distorted, has incorrect timing, or is logically contradictory. Control stability involves limiting the slope and amplitude of control command adjustments to avoid sudden increases or decreases in power and torque, prevent frequent alternating adjustments by multiple mechanisms, and ensure smooth power output without oscillations or driving jerks. Control real-time performance ensures that parameter processing, command generation, and issuance delays meet the vehicle control cycle, and can synchronously respond to instantaneous changes in gradient and load without control lag. Control safety ensures that battery temperature, charging and discharging, motor speed and torque, and electrical equipment voltage and current are all within rated thresholds, avoiding the risks of overload, overcharging, over-discharging, and over-temperature operation.

[0040] After receiving the control command, the physical power unit controls the power output and energy distribution of each actuator according to the control command.

[0041] like Figure 2As shown, the vehicle control unit (VCU) is located in the tractor's onboard control system. It receives real-time operating status parameters from the physical power unit and generates corresponding control commands to achieve coordinated control of the various actuators within the physical power unit. The VCU includes at least a power management module, a travel / operation control module, and a braking and safety control module.

[0042] The power management module is used to coordinate the flow of oil and electricity between the range extender's power generation and the battery's charging and discharging based on the target and constraint parameters and the vehicle's energy allocation priority issued by the digital twin unit, combined with the economic operating range of the range extender and the safe charging and discharging boundary of the battery. The range extender prioritizes supplying working power to the track drive motor and PTO work motor, and the remaining power is used to charge the battery system through the power conversion device. The battery system releases power under low load conditions and provides power in coordination under heavy load / climbing conditions, realizing real-time matching and closed-loop flow of range extender power generation, battery charging and discharging and motor energy consumption. It manages the energy flow of the range extender, battery system and power conversion device (i.e. the left and right drive motors corresponding to the left and right tracks), collects operating status parameters and sends input data to the digital twin unit, and receives operating status feedback to coordinate power and energy scheduling.

[0043] The walking / operation control module is responsible for the vehicle's walking drive and field operation power output; the walking / operation control module is used for the walking and driving control of the left and right drive motors and the power output of the PTO motor for the work load. It executes the walking and operation functions of the whole machine according to the instructions of the vehicle controller, and matches the terrain and work load conditions.

[0044] The braking and safety control module is responsible for the overall machine's operational safety constraints, operational condition protection, and anomaly management. Relying on terrain and load data from the operational condition acquisition device, and in conjunction with the operating status parameters of each actuator, the braking and safety control module performs safety limit control (achieved through the execution units of each actuator, with each execution unit corresponding to...). Figure 2 The "braking" component ensures that the range extender, battery, and motor operate under safe conditions, serving as a safety feedback and braking protection execution unit for the control execution effect.

[0045] The digital twin platform interacts with the VCU and physical power unit through a communication and data interaction layer. This layer is used to evaluate and optimize the power unit based on synchronized operational status data and output optimized control parameters (including target parameters and constraint parameters) to the VCU. The digital twin platform is equipped with a working condition prediction and evaluation module and an optimization and fault-tolerant decision-making module. The working condition prediction and evaluation module constructs a working condition digital model based on the temporal relationship between terrain data and operational load data; it performs working condition simulations based on slope information, surface adhesion indicators, and equivalent resistance coefficients, combined with temporal correlation patterns; it inputs real-time working condition data and extracts predicted working condition data within a preset prediction window; it jointly fits the power model and the working condition digital model to conduct simulation analysis, completing a quantitative assessment of the overall machine power demand, energy flow of each actuator, and power distribution, thus achieving prediction of working conditions in hilly and mountainous areas and assessment of the overall machine power load. The optimization and fault-tolerant decision-making module obtains a fitted dynamic model by correcting the virtual dynamic model based on real-time operating state parameters, and completes the self-correction of the model's dynamic accuracy. It integrates real-time operating condition data and predicted operating condition data for joint simulation calculation, and processes the simulation analysis results into target parameters and constraint parameters. It standardizes and encapsulates the decision parameters and sends them to the data interaction unit to provide optimized benchmark parameters for vehicle control. At the same time, relying on the real-time model correction mechanism and operating condition timing prediction logic, it provides upfront fault-tolerant data support for subsequent parameter anomalies and communication disturbances, ensuring that the decision output is reliable and usable.

[0046] The communication and data interaction layer is used to realize bidirectional data transmission between the physical power unit, the VCU vehicle controller, and the digital twin platform, supporting the transmission of various data. Operating status parameters are acquired via the CAN bus, the physical power unit controls the operation of each actuator via an interface, and interaction with the digital twin platform is achieved via Ethernet.

[0047] The digital twin unit constrains the value range of the surface adhesion index through slope information; it correlates the variation law of the equivalent resistance coefficient by combining the surface adhesion index with the pre-set inherent characteristics of the field soil; and it temporally couples and correlates the topographic data of various regions. The working condition digital model, based on the coupled topographic data and operational load data, quantifies the relationship between slope changes, surface adhesion conditions, soil resistance, and operational load on the overall machine power demand during hilly and mountainous operations. Using slope information, equivalent resistance coefficient, and adhesion index as input parameters, the working condition digital model quantifies the contribution ratio of slope, surface adhesion, and soil resistance to driving resistance through multi-dimensional topographic coupling and correlation calculations and load temporal matching analysis. It then performs weighted fitting based on real-time operational load temporal characteristics to obtain the equivalent resistance representation quantity and load demand representation quantity under the corresponding operational scenario. The fitted power model is obtained by correcting the virtual power model with the equivalent resistance representation quantity and load demand representation quantity.

[0048] Among them, the operating environment parameters can be obtained through measurement, online estimation, or operating condition sequence playback using dedicated terrain and load sensors such as slope sensors, surface adhesion detection sensors, and resistance sensors. These parameters serve as state variables in the digital twin model for prediction and optimization calculations. Sensors collect data on the status of the power battery, the operating status of the drive motor, the operating status of the range extender, the operating load status, and the overall vehicle travel status. Combined with control commands, the current operating conditions of each actuator are identified and transmitted to the digital twin platform via the CAN bus.

[0049] Specifically, online estimation relies on the real-time operating status of tractors and historical time-series data of terrain load. Through iterative deduction of the virtual dynamic model, parameters such as terrain resistance and adhesion index, which cannot be directly measured by sensors, are supplemented. The working condition sequence playback retrieves historical typical slope, soil resistance, and working load time-series working condition segments in the field, and reproduces the working condition changes according to the actual operation process. The parameters obtained by both methods are simultaneously incorporated into the state variables of the virtual dynamic model and participate in working condition deduction, power demand prediction, and model optimization calculation.

[0050] The digital twin unit contains state variables to characterize the operating status and parameter characteristics of the virtual power model and the operating condition digital model. The physical power unit extracts key state parameters from the operating status parameters and operating condition data, and transmits these key state parameters to the digital twin unit in real time via the tractor's CAN bus. After receiving the key state parameters, the digital twin unit uses them to correct the state variables of the virtual power model and the operating condition digital model. Based on the corrected state variables, the virtual power model derives the predicted operating results corresponding to power demand, energy flow, and power distribution.

[0051] The physical power unit collects the actual operating results of each actuator of the tractor under actual working conditions and sends them to the digital twin unit. The digital twin unit compares the predicted operating results derived from the virtual power model with the actual operating results collected by the physical power unit to obtain the operating deviation. Based on the hardware structure, parameter characteristics, and operating rules of each actuator on which the virtual power model is based, the digital twin unit extracts the model structure parameters, characteristic fitting parameters, and operating condition correlation parameters in the process of fitting the virtual power model to form the fitted power model, and uses them as internal key parameters. The digital twin unit performs online calibration of the internal key parameters built into the virtual power model based on the operating deviation, so that the virtual power model and the operating condition digital model can continuously iterate and evolve as the tractor equipment status and operating environment change.

[0052] Specifically, the digital twin unit relies on the hardware and operating conditions of each actuator to dynamically fine-tune key internal parameters such as model structure parameters, characteristic fitting parameters, and operating condition correlation parameters, and corrects model simulation errors in a closed loop, driving the virtual power model and the operating condition digital model to iterate and update synchronously as the equipment ages and the environment changes.

[0053] The digital twin unit constructs a virtual power model and a working condition digital model using a reduced-order model or a proxy model, enabling it to operate stably on edge computing devices or upper-level computing units at a timescale slower than the vehicle control cycle, providing reliable optimization decision support for the vehicle controller. The digital twin model sets its computation cycle according to the timing of the tractor's overall vehicle control cycle, which is slower than the timing of the tractor's overall vehicle control cycle. Within the computation cycle, the digital twin unit synchronously updates the state variables and internal key parameters of the virtual power model and the working condition digital model. Based on the updated state variables and calibrated internal key parameters, it generates optimization decision-related parameters, including target parameters and constraint parameters, and sends these parameters to the decision control unit.

[0054] The edge computing device is deployed locally on the tractor, responsible for real-time collection of vehicle operation and working condition sensor data. It performs data preprocessing, time-series alignment, anomaly identification, and preliminary computing calculations on-site, and distributes basic control parameters locally, supporting vehicle movement, operation, and safety control with low latency. The upper-level computing unit communicates with the on-board edge computing device, aggregates preprocessed data from the edge, performs lightweight model simulation, local parameter optimization, and decision calculations, and forwards standardized business data upwards. The digital twin platform is deployed on a remote server, receiving full-volume time-series data uploaded by the upper-level computing unit, completing full-dimensional twin modeling, online parameter calibration, working condition prediction, and energy strategy optimization, and then distributing target parameters and constraint parameters to the upper-level computing unit for execution via the edge computing device.

[0055] The decision control unit performs boundary constraint checks, rate of change constraint judgments, and safety condition verifications on target parameters and constraint parameters based on preset verification conditions (including verification of parameter value abrupt changes exceeding limits, logical inconsistencies, data incompleteness, and timing frame anomalies). When the target parameters and constraint parameters meet the preset verification conditions, the decision control unit sets the target parameters and constraint parameters as the vehicle target control parameters and generates control commands based on the vehicle target control parameters. When the target parameters and constraint parameters do not meet the preset verification conditions or when complete target parameters and constraint parameters are not received within the preset timeout window, the preset local control strategy is invoked to generate backup control parameters, and control commands are generated based on the backup control parameters.

[0056] The vehicle controller generates start / stop and target speed commands for the range extender, battery charging / discharging power commands, and left and right drive motor torque commands based on backup control parameters, and sends them to the corresponding execution units to achieve coordinated control of each actuator. Simultaneously, it collects the execution results of each actuator and feeds them back to the digital twin platform through a data interaction unit to update the parameters of the virtual power model, enabling continuous correction and rolling optimization of the virtual power model's operating characteristics.

[0057] The control object of the digital twin platform is the key controllable operating parameters in the real-time power control system of the range-extended tracked tractor based on digital twin. The control object is sent to the vehicle controller via CAN bus in the form of target value, constraint parameter and weight parameter, and the vehicle controller completes the integration and execution.

[0058] The weight parameters are based on the power response sensitivity of each actuator, the degree of influence of operating conditions, the safety margin of components and energy consumption priority, combined with the slope resistance of hilly terrain and the fluctuation characteristics of operating load, and the basic weights are calibrated offline. Then, the real-time operating condition adaptability, battery energy storage and temperature constraint status are identified online through the digital twin platform, and the allocation ratio of each parameter is dynamically corrected. Finally, weight parameters that can be distributed are generated for the multi-objective control fusion scheduling of the vehicle controller.

[0059] The controlled objects include at least the range extender operating parameters, the power battery operating parameters, the left and right track drive motor operating parameters, the PTO work motor operating parameters, and the overall constraint parameters.

[0060] Specifically, the operating parameters of the range extender include the range extender start-stop status, the target operating speed and target power generation set by the administrator based on the rated speed of the range extender engine and the economic speed range.

[0061] Specifically, the operating parameters of the power battery include the upper limit of allowable charge and discharge power, the target state of charge range, and the energy allocation priority. The energy allocation priority is set by the administrator based on the overall vehicle operating mode, real-time operating load, power response characteristics of each actuator, battery status, and component safety constraints.

[0062] Specifically, the operating parameters of the left and right track drive motors include the target output torque, speed limit, and power distribution ratio of each drive motor. The power distribution ratio is determined by the administrator based on the ratio of the target output torque on the left and right sides, the constraint range of the speed limit, and the difference between the surface adhesion coefficient and terrain resistance, so that the torque and speed on both sides are distributed proportionally within the safety limits, achieving smooth straight-line movement, controllable steering, no slippage, and no exceeding of limits.

[0063] Specifically, the operating parameters of the PTO work motor include the work power requirement and the power allocation weight. The work power requirement is determined by the real-time work load and the working status of the equipment. The power allocation weight is determined by the administrator based on the size of the work power requirement, the battery energy reserve, the total power limit of the vehicle, and the work priority level. The greater the requirement, the higher the weight, to ensure a stable supply of work power without exceeding the safety constraints of the vehicle.

[0064] Specifically, the overall constraint parameters include the system's total power limit, dynamic safety margin, and energy management constraints. The system's total power limit is determined by the administrator based on the sum of the rated power of each power component and the bench calibration safety threshold. The dynamic safety margin is reserved based on the overload capacity of the components and the fluctuation range of operating conditions. The energy management constraints are determined comprehensively based on the battery safety boundary, the stable operating range of the range extender, and the vehicle control specifications.

[0065] This embodiment also includes a power control method for range-extended tracked tractors based on digital twins, such as... Figure 3 As shown, the method includes the following steps: S1. By collecting operating status parameters such as the status of the power battery, the operating status of the drive motor, the working status of the range extender, the working load status, and the vehicle's travel status through sensors arranged in the physical power unit, the operating conditions of the vehicle are identified in combination with the current work instructions.

[0066] S2. Synchronize the operating status parameters to the digital twin platform, drive the virtual power model to update the operating status, and evaluate the vehicle power demand based on the current operating conditions.

[0067] S3. Based on the power demand assessment results, coordinate and match the power distribution relationship between the power battery system, range extender, left and right track drive motors and PTO working motor, and generate corresponding control parameters in accordance with the range extender's economic power generation and battery priority power supply exclusive control logic.

[0068] The administrator sets up a dedicated control logic for the range extender's economical power generation based on the economical speed range and the maximum efficiency MAP curve calibrated on the range extender test bench. This logic only activates when the battery's SOC is below a preset lower limit or when the vehicle is under heavy load, ensuring it operates at the point of lowest fuel consumption and highest power generation efficiency, thus avoiding inefficient fuel consumption. A dedicated control logic prioritizing battery power supply is also set up based on the battery's safe SOC range and real-time power demand. Under light load, idling, and medium load conditions, the battery system prioritizes powering the motor, temporarily deactivating the range extender to maximize electrical energy utilization and reduce fuel consumption. The digital twin unit anticipates operating conditions and power demands, and the decision control unit allocates energy according to the rule of prioritizing battery power supply followed by economical range extender charging. The power management module strictly executes this logic to achieve optimal scheduling of hybrid and electric energy.

[0069] S4. Determine whether to enable the optimized control parameters output by the digital twin platform based on the operating status parameters; if digital twin optimization is enabled, use the optimized control parameters output by the digital twin platform; if not enabled, use the local control parameters inside the VCU vehicle controller.

[0070] S5, the VCU vehicle controller generates control commands based on the selected control parameters and applies them to each actuator in the physical power unit, while simultaneously acquiring the operating status parameters after execution.

[0071] S6. Determine whether the operating deviation exceeds the preset threshold based on the acquired operating status parameters; if the operating deviation does not exceed the threshold, maintain the current control state; if the operating deviation exceeds the threshold, return to step S2 to re-evaluate and match the power demand.

[0072] The digital twin unit extracts the variation and periodic characteristics of historically acquired operating status parameters and working condition data as verification features, and extracts the variation and periodic characteristics of the working condition data and operating status parameters received in the current calculation cycle as comparison features. When the difference between the comparison features and the verification features is greater than the preset verification difference, the digital twin unit combines the historically acquired operating status parameters and working condition data with the virtual dynamic model and the working condition digital model to verify and compensate for the state variables and internal key parameters of the virtual dynamic model and the working condition digital model in the current calculation cycle. The compensated state variables and internal key parameters are then sent to the decision control unit as substitute parameters. The decision control unit generates control commands based on the received substitute parameters.

[0073] like Figure 4 As shown, under normal operating conditions, the real-time operating status parameters of the physical power unit are transmitted directly to the VCU vehicle controller as real-time operating data to support real-time vehicle control. Simultaneously, the operating data is synchronized to the digital twin platform via the communication and data interaction layer to drive the virtual power model for status updates and evaluation. Based on the evaluation results, the digital twin platform outputs optimized control parameters to the VCU vehicle controller. The VCU vehicle controller then generates control commands based on these optimized control parameters and applies them to the vehicle subsystems (including the VCU vehicle controller and various actuators).

[0074] In abnormal operating conditions such as sensor malfunctions or missing operational data, the digital twin platform estimates and compensates for missing or abnormal parameters based on historical operating data and a virtual power model, and outputs alternative control parameters to the VCU vehicle controller. These alternative control parameters include the range extender's generator power and start / stop timing. The VCU vehicle controller continues to execute control based on these alternative control parameters, thereby ensuring the continuity and stability of vehicle control under abnormal operating conditions.

[0075] Example 2

[0076] The only difference between this embodiment and Embodiment 1 is that the physical power unit also includes temperature sensors for monitoring the battery system temperature and ambient temperature. These temperature sensors are installed in the battery compartment and on the outside of the tractor body. The physical power unit sends the battery system temperature and ambient temperature to the digital twin unit via a data interaction unit. The collected temperature data (battery system temperature and ambient temperature) is synchronized to the digital twin unit after being time-aligned by the data interaction unit. The temperature data provides boundary constraints for subsequent overall energy consumption statistics, battery energy inventory correction, and operating efficiency prediction.

[0077] The digital twin unit defines the total energy consumed by various actuators, such as the range extender, track drive motor, and PTO work motor, during tractor operation as the overall machine energy consumption. The range extender, as an on-board power generation actuator, continuously provides the tractor with energy converted from gasoline to electricity. The remaining usable energy that the battery system can continuously output within the safe operating range is defined as the battery energy reserve. After receiving real-time data on battery system temperature and ambient temperature, the digital twin unit, based on the inherent operating mechanism of the virtual power model, simulates the inherent correlation between the overall machine energy flow path, energy conversion loss, and battery system operating temperature under different operating temperature conditions. At the same time, the administrator pre-sets the battery safety temperature threshold range based on the battery characteristics and the boundary of the hilly operating environment, limiting the allowable operating power range of the battery system participating in overall machine energy conversion under different temperature conditions, and predicts the changes in overall machine operating efficiency in subsequent short-term operating periods based on the evolution characteristics of the operating conditions.

[0078] The digital twin unit uses ambient temperature, real-time operating efficiency, energy conversion duration, and battery system temperature changes caused by efficiency fluctuations as related parameters. It constructs a dynamic battery temperature model based on the temporal coupling relationship and operating condition matching characteristics between these parameters. Through layered logic built into the model, it sets battery temperature constraints to routinely manage key operating parameters such as the battery's charge / discharge range and power output amplitude. This dynamic battery temperature model builds upon the correlation between temperature and energy conversion from previous models, providing a stable simulation platform for subsequent energy conversion condition prediction and battery energy inventory correction. The temperature constraints output by the model can simultaneously exert unified constraints on the range extender's power generation, track drive motor power allocation, and PTO operating power weights, achieving collaborative constraint control of multiple actuators within a single model.

[0079] The digital twin unit, based on the constructed dynamic battery temperature model, predicts the evolution of the overall machine's energy conversion conditions and the temperature rise and fall trends of the battery system during subsequent operation periods. It also performs operational condition logic matching simulations based on the tractor's real-time walking power requirements and field operation load demands, matching the range extender's economic power generation range with the battery's safe charging and discharging range. Specifically, the administrator sets the economic power generation range based on the range extender's efficiency MAP curve calibrated on the test bench and the optimal economic speed or power range, and sets the battery's safe charging and discharging range based on the battery manufacturer's specifications, charge / discharge rate, temperature characteristics, and safe lifespan requirements. The digital twin unit links and matches these two parameters, ensuring that the range extender's power output always matches the battery's safe charging and discharging range, achieving efficient hybrid operation.

[0080] Based on the simulation results, an adaptive energy conversion strategy is formulated to match the current temperature and load conditions. Through preset strategy logic, the timing of battery system discharge and the timing and duration of the range extender charging the battery system are precisely controlled. This adaptive energy conversion strategy integrates multiple characteristic information from previous temperature correlation patterns, model constraints, and overall energy consumption benchmarks. The strategy output can be sent to the decision control unit to be converted into vehicle control commands, and can also provide feedback to iteratively correct the internal characteristic parameters of the virtual power model, while providing timing benchmark support for battery energy inventory correction in the next calculation cycle.

[0081] The digital twin unit summarizes and fits the fluctuation patterns of overall energy consumption under different operating environment temperatures. Combining the real-time operating environment temperature and the real-time battery system temperature, and relying on the established temperature and energy consumption correlation mapping features, it dynamically corrects the real-time identified battery energy inventory for different actual operating conditions such as hilly slope undulations and sudden changes in operating load. The corrected battery energy inventory not only completes the real-time update of its own state parameters, but also serves as the core basis for determining the start-stop status of the range extender in the next cycle, setting the upper limit of the allowable charging and discharging power of the power battery, and configuring the energy allocation priority and power allocation weight of each actuator. This realizes the reverse empowerment of front-end multi-parameter optimization decision-making by correcting the end-of-line inventory, forming a closed-loop coupled optimization mechanism throughout the entire process.

[0082] In this embodiment, the battery system temperature and ambient temperature data collected in real time by the physical power unit are transmitted to the digital twin unit via the data interaction unit. This provides basic input parameters for building the battery temperature dynamic model, as well as environmental boundary basis for predicting the overall machine's operating efficiency. Simultaneously, it provides temperature benchmark support for battery operating parameter constraints and adaptive energy conversion strategy formulation, and synchronously adapts to the range extender's hybrid energy scheduling rules (set by the administrator, with battery priority power supply and economical range extender power generation as the core, combining operating conditions and temperature to coordinate hybrid energy allocation, start-up, shutdown, and charging / discharging). The temperature and energy conversion correlation established by the digital twin unit provides operating condition coupling basis for building the battery temperature dynamic model, with the model being set in layers. Temperature constraints, in turn, limit the power adjustment range of the energy conversion strategy. The trend prediction results of energy conversion conditions not only support the logical generation of adaptive energy conversion strategies, but also provide a time-series change reference for the dynamic correction of battery energy inventory. The corrected and updated battery energy inventory parameters can also participate in the prediction of operating conditions, fine-tuning of model parameters, and adjustment of energy allocation weights in the next calculation cycle. Each link relies on data correlation and operating condition mapping relationship to connect and link in both directions. The characteristic patterns formed in the preceding links are continuously integrated into the entire process of subsequent modeling, trend prediction, parameter control, and inventory correction. The operating status and deviation characteristics generated in the subsequent links are also fed back layer by layer to iteratively correct the internal matching characteristics of the previous temperature correlation patterns and the battery temperature dynamic model.

[0083] Without adding extra hardware configuration and complex computing logic, this embodiment can achieve adaptive temperature control of the tractor power system under different ambient temperatures and different load conditions, efficient energy flow and precise parameter matching, effectively improving the smoothness of the vehicle's power output, battery operation safety and overall energy utilization efficiency in hilly and mountainous areas with varying temperatures and loads. At the same time, it can still ensure the continuity of vehicle power control and operational stability in scenarios with small fluctuations in sensor data and gradual changes in operating environment.

[0084] The above are merely embodiments of this application. This invention is not limited to the field covered by this embodiment. Commonly known structures and characteristics in the solution are not described in detail here. Those skilled in the art are aware of all common technical knowledge in the field prior to the application date or priority date, are able to access all existing technologies in that field, and have the ability to apply conventional experimental methods prior to that date. Those skilled in the art can, based on the guidance provided in this application, improve and implement this solution in combination with their own capabilities. Some typical known structures or methods should not be obstacles for those skilled in the art to implement this application. It should be noted that those skilled in the art can make several modifications and improvements without departing from the structure of this application. These should also be considered within the scope of protection of this application, and will not affect the effectiveness of the implementation of this application or the practicality of the patent. The scope of protection claimed in this application should be determined by the content of its claims. The specific embodiments described in the specification can be used to interpret the content of the claims.

Claims

1. A real-time power control system for a range-extended tracked tractor based on digital twins, characterized in that, include: The physical power unit includes various actuators mounted on the tractor body, each actuator including a range extender, a battery system, a PTO motor, a power conversion device, and a track drive motor; it is used to collect the operating status parameters of each actuator in real time; it also includes a working condition acquisition device for collecting working condition data in real time, the working condition data including terrain data and work load data; The data interaction unit is used to acquire and time-series align operating condition data and running status parameters; The digital twin unit constructs a virtual power model based on the hardware structure, parameter characteristics, and operating rules of each actuator, and constructs a working condition digital model based on the temporal relationship of terrain data and work load data. It acquires real-time operating status parameters and real-time working condition data according to the acquisition time, and corrects the virtual power model based on the real-time operating status parameters to obtain a fitted power model. It inputs the real-time working condition data into the working condition digital model and obtains the working condition data within a preset prediction window as predicted working condition data. Based on the real-time working condition data and predicted working condition data, it performs simulation analysis by fitting the power model and the working condition digital model to obtain the tractor's power demand, energy flow between actuators, and power distribution. The obtained data is then processed into target parameters and constraint parameters, which are sent to the data interaction unit. The decision control unit receives the target parameters and constraint parameters sent by the data interaction unit, and performs fusion processing with real-time operating data and preset safety constraints to obtain control commands; the control commands are then sent to the physical power unit through the data interaction unit. After receiving the control command, the physical power unit controls the power output and energy distribution of each actuator according to the control command.

2. The real-time power control system for range-extended tracked tractors based on digital twins according to claim 1, characterized in that: The digital twin unit is also used to construct a range extender generator model, a power battery model, a drive motor model, and a power distribution model based on the hardware structure of each actuator. The range extender generator model is used to characterize the changes in the generator's power output and efficiency under different operating conditions. The power battery model is used to characterize the changes in the battery's state of charge, charging and discharging power limitations, and dynamic response characteristics. The drive motor model is used to characterize the torque output characteristics and efficiency distribution of the track drive motor. The power distribution model is used to characterize the power coordination relationship between the power sources of each actuator and the system constraints. The range extender generator model, power battery model, drive motor model, and power distribution model are integrated to form a virtual power model.

3. The real-time power control system for range-extended tracked tractors based on digital twins according to claim 1, characterized in that: The terrain data includes slope information, surface adhesion indicators, and equivalent resistance coefficient; The value range of the surface adhesion index is constrained by the slope information; the variation law of the equivalent resistance coefficient is correlated by combining the surface adhesion index with the pre-set inherent characteristics of the field soil; and the topographic data of various regions are coupled and correlated in time series. The working condition digital model, based on coupled terrain data and operational load data, quantifies the effects of slope changes, surface adhesion conditions, soil resistance, and operational load on the overall power demand of the machine during hilly and mountainous operations. The working condition digital model uses slope information, equivalent resistance coefficient, and adhesion index as input parameters to analyze and obtain the equivalent resistance representation quantity and load demand representation quantity under the corresponding operational scenario. The virtual power model, corrected by the equivalent resistance representation quantity and load demand representation quantity, yields the fitted power model.

4. The real-time power control system for range-extended tracked tractors based on digital twins according to claim 1, characterized in that: The digital twin unit includes state variables to characterize the operating status and parameter characteristics of the virtual dynamic model and the digital working condition model; The physical power unit extracts key state parameters from operating status parameters and working condition data, and transmits these key state parameters to the digital twin unit in real time via the tractor's CAN bus. After receiving the key state parameters, the digital twin unit uses them to correct the state variables of the virtual power model and the working condition digital model. Based on the corrected state variables, the virtual power model deduces the predicted operating results corresponding to power demand, energy flow, and power distribution. The physical power unit collects the actual operating results of each actuator of the tractor under actual working conditions and sends them to the digital twin unit; The predicted operating results obtained from the virtual dynamic model are compared with the actual operating results collected by the physical dynamic unit to obtain the operating deviation; The digital twin unit extracts model structure parameters, characteristic fitting parameters, and operating condition correlation parameters from the process of fitting the virtual power model to form the fitted power model based on the hardware structure, parameter characteristics, and operating rules of each actuator on which the virtual power model is constructed, and uses them as internal key parameters. The digital twin unit performs online calibration of the key internal parameters built into the virtual dynamic model based on operational deviations.

5. The real-time power control system for range-extended tracked tractors based on digital twins according to claim 4, characterized in that: The digital twin unit constructs a virtual power model and a working condition digital model using a reduced-order model or a proxy model. The digital twin model sets its calculation cycle according to the timing of the tractor's overall control cycle, and the calculation cycle is slower than the timing of the tractor's overall control cycle. Within the calculation cycle, the digital twin unit synchronously updates the state variables and internal key parameters of the virtual power model and the working condition digital model, and generates optimization decision-related parameters, including target parameters and constraint parameters, based on the updated state variables and calibrated internal key parameters. The optimization decision-related parameters are then sent to the decision control unit.

6. The real-time power control system for range-extended tracked tractors based on digital twins according to claim 1, characterized in that: The decision control unit performs boundary constraint checks, rate of change constraint judgments, and safety condition verifications on the target parameters and constraint parameters according to preset verification conditions. When the target parameters and constraint parameters meet the preset verification conditions, the decision control unit sets the target parameters and constraint parameters as the vehicle target control parameters and generates control commands based on the vehicle target control parameters. If the target parameters and constraint parameters do not meet the preset verification conditions or if the target parameters and constraint parameters are not received within the preset timeout window, the preset local control strategy is invoked to generate backup control parameters, and control commands are generated based on the backup control parameters.

7. The real-time power control system for range-extended tracked tractors based on digital twins according to claim 5, characterized in that: The digital twin unit extracts the change characteristics and periodic characteristics of historically collected operating status parameters and operating condition data as verification characteristics, and extracts the change characteristics and periodic characteristics of the operating condition data and operating status parameters received in the current operation cycle as comparison characteristics. When the difference between the comparison characteristics and the verification characteristics is greater than the preset verification difference, the digital twin unit combines the historically collected operating status parameters and operating condition data with the virtual power model and the operating condition digital model to verify and compensate the state variables and internal key parameters of the virtual power model and the operating condition digital model in the current operation cycle. The compensated state variables and internal key parameters are then used as substitute parameters and sent to the decision control unit. The decision control unit generates control commands based on the received alternative parameters.

8. The real-time power control system for range-extended tracked tractors based on digital twins according to claim 2, characterized in that: The physical power unit also includes temperature sensors that monitor the battery system temperature and ambient temperature. The physical power unit sends the battery system temperature and ambient temperature to the digital twin unit through the data interaction unit. The digital twin unit takes the total energy consumed by each actuator during the operation of the tractor as the energy consumption of the whole machine, and the remaining energy that the battery system can currently output as the battery energy reserve. The digital twin unit simulates the correlation between the overall energy conversion process and the battery system temperature under different operating ambient temperatures based on the received battery system temperature and ambient temperature; it limits the battery system energy conversion power according to the preset battery safety temperature threshold and predicts the overall operating efficiency through real-time operating ambient temperature. The digital twin unit uses the operating environment temperature, overall machine operating efficiency, energy conversion time, and battery system temperature changes caused by efficiency changes as parameters to build a dynamic battery temperature model, and controls battery operating parameters by setting battery temperature constraints. The digital twin unit combines a dynamic battery temperature model to predict energy conversion conditions and battery system temperature change trends. It performs logical calculations based on the real-time power demand of the tractor, formulates an adaptive energy conversion strategy based on the calculation results, and controls the timing of external discharge and recharging of the battery system through the energy conversion strategy.

9. The real-time power control system for range-extended tracked tractors based on digital twins according to claim 8, characterized in that: The digital twin unit fits the overall energy consumption variation pattern of the machine under different operating environment temperatures. Combining the real-time operating environment temperature and the real-time battery system temperature, it dynamically corrects the battery energy storage under different operating conditions based on the overall energy consumption variation pattern.

10. A real-time power control method for range-extended tracked tractors based on digital twins, characterized in that, The real-time power control system for range-extended tracked tractors based on digital twins, as described in any one of claims 1-9, was used.

Images