Pure electric drive bulldozer power battery state estimation and cooperative control method and system

Abstract

The present disclosure provides a pure electric drive bulldozer power battery state estimation and collaborative control method and system, relating to the technical field of pure electric drive control, comprising: constructing a whole machine multi-physical field coupling model; based on the whole machine multi-physical field coupling model, considering vibration and temperature factors, constructing a power battery electric-thermal coupling model; based on the power battery electric-thermal coupling model, analyzing the key factors affecting the accuracy of battery SOH estimation, and inputting into the gating cycle unit network to output the preliminary estimation result of SOH; correcting the preliminary estimation result of SOH to obtain the final estimation of battery SOH; setting the battery power boundary, constructing the objective function with the lowest battery power consumption and equivalent battery life attenuation cost, using the model predictive control method and combining the generalized minimal residual numerical algorithm to quickly solve, realizing the collaborative optimization of motor torque and battery power. The present disclosure improves the adaptability of battery state estimation accuracy and control strategy.

Description

Technical Field

[0001] This disclosure relates to the field of pure electric drive control technology, specifically to a method and system for estimating the state of the power battery and coordinating control of a pure electric drive bulldozer. Background Technology

[0002] The statements in this section are merely background information relating to this disclosure and do not necessarily constitute prior art.

[0003] To address energy shortages and environmental pollution, developing new energy vehicles has become a strategic choice for the automotive industry. The transformation of traditional construction machinery to electric power is a crucial way to solve energy security issues. Traditional construction machinery primarily uses diesel engines as its power source, characterized by high energy consumption and emissions, with diesel engine efficiency only reaching 30%-40%. Pure electric construction machinery, as an important component of new energy construction machinery, offers advantages such as low energy consumption and zero emissions, reducing overall operating costs by over 60%. Its key technologies are accurate estimation of the battery's state of health (SOH) and vehicle control strategies. Accurate SOH estimation not only meets the requirements for safe and efficient battery use but also provides crucial state information for vehicle control strategies, directly determining battery lifespan and overall machine economy.

[0004] However, due to the harsh working environment, high impact, complex and drastic changes in operating conditions of construction machinery, accurate estimation of battery status is extremely challenging.

[0005] Existing methods include Equivalent Circuit Model (ECM)-based methods, which simulate battery dynamics using components such as resistors and capacitors and combine them with Kalman filtering (EKF / UKF) for SOH estimation. Electrochemical model-based methods, based on internal battery chemical reactions (such as the P2D model), characterize SOH through parameter degradation. However, these methods have poor environmental adaptability; the severe vibrations and extreme temperatures (-30℃ to 60℃) of construction machinery cause model parameter drift and decreased accuracy. Electrochemical models require extensive real-time calculations, making them difficult to embed into vehicle controllers. Furthermore, model parameters need to be calibrated for different operating conditions, and the variable operating conditions of construction machinery lead to insufficient generalization. Summary of the Invention

[0006] To address the aforementioned issues, this disclosure proposes a method and system for state estimation and coordinated control of the power battery in a pure electric bulldozer. It establishes a load spectrum and multi-physics coupling model for the entire pure electric bulldozer, constructs an electro-thermal coupling model for the battery considering vibration and temperature factors, and integrates a mechanistic model and a data-driven method to achieve highly accurate estimation of the power battery's state of equilibrium (SOH). A multi-objective optimization control problem weighted by energy consumption and lifespan is constructed, and a model predictive control (MPC) method combined with a generalized minimum residual numerical algorithm is used for rapid solution to achieve coordinated optimization of motor torque and battery power.

[0007] According to some embodiments, the present disclosure adopts the following technical solutions: A method for state estimation and coordinated control of the power battery in a pure electric bulldozer, including: Obtain typical working condition data of the pure electric bulldozer, identify the typical working conditions, and construct a multi-physics coupling model of the whole machine based on the identification results; Based on the multi-physics coupling model of the whole machine, and considering vibration and temperature factors, an electro-thermal coupling model of the power battery is constructed. Based on the electro-thermal coupling model of the power battery, the key factors affecting the accuracy of battery SOH estimation are analyzed. The key factors are input into the gated recurrent unit network, and the preliminary SOH estimation result is output. The adaptive extended Kalman filter algorithm is used to correct the preliminary SOH estimation result to obtain the final SOH estimate of the battery. Based on the final estimate of battery SOH, the battery power boundary is set. Considering the influence of the battery power boundary and battery SOH on battery life, a battery life degradation model is established. The objective function is constructed with the minimum consumed battery power and the equivalent battery life degradation cost. The model predictive control method is adopted and combined with the generalized minimum residual numerical algorithm to solve the problem quickly, so as to achieve the coordinated optimization of motor torque and battery power.

[0008] According to some embodiments, the present disclosure adopts the following technical solutions: A state estimation and cooperative control system for the power battery of a pure electric bulldozer, including: The whole vehicle model building module is used to acquire typical working condition data of the pure electric bulldozer, identify the typical working conditions, and build a multi-physics coupling model of the whole machine based on the identification results. The electrothermal coupling model construction module is used to construct an electrothermal coupling model of the power battery based on the multi-physics coupling model of the whole machine, taking into account vibration and temperature factors. The battery estimation module is used to analyze the key factors affecting the accuracy of battery SOH estimation based on the power battery electro-thermal coupling model. The key factors are input into the gated recurrent unit network and the output is a preliminary SOH estimation result. The preliminary SOH estimation result is corrected by an adaptive extended Kalman filter algorithm to obtain the final estimate of battery SOH. The collaborative optimization module is used to set the battery power boundary based on the final estimate of the battery SOH. It considers the impact of the battery power boundary and the battery SOH on the battery life, establishes a battery life degradation model, constructs the objective function with the minimum consumed battery power and the equivalent battery life degradation cost, and uses model predictive control method combined with generalized minimum residual numerical algorithm to solve it quickly, so as to achieve collaborative optimization of motor torque and battery power.

[0009] According to some embodiments, the present disclosure adopts the following technical solutions: A computer program product includes a computer program that, when executed by a processor, implements the method for estimating the state of the power battery and coordinating control of a pure electric bulldozer.

[0010] According to some embodiments, the present disclosure adopts the following technical solutions: A non-transitory computer-readable storage medium is provided for storing computer instructions, which, when executed by a processor, implement the aforementioned method for estimating and coordinating the state of the power battery of a pure electric bulldozer.

[0011] According to some embodiments, the present disclosure adopts the following technical solutions: An electronic device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, the computer program is stored in the memory, and when the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to implement the method for estimating and coordinating the state of the power battery of a pure electric bulldozer.

[0012] Compared with the prior art, the beneficial effects of this disclosure are as follows: This disclosure presents a method for estimating the state of power (SOH) of a pure electric bulldozer's power battery and for coordinated control. Addressing the challenges of accurately assessing battery state and poor control strategy adaptability in complex operating conditions of construction machinery, this method constructs a load spectrum and multi-physics coupling model for the entire pure electric bulldozer. It establishes a battery electro-thermal coupling model considering vibration and temperature factors, integrates a mechanistic model and a data-driven approach, and utilizes the GRU-AEKF joint algorithm to achieve highly accurate SOH estimation of the power battery. Based on this foundation, considering the SOP boundary and battery life degradation, a multi-objective optimization control problem weighted by energy consumption and lifespan is constructed. Model Predictive Control (MPC) is employed, combined with a generalized minimum residual numerical algorithm for rapid solution, to achieve coordinated optimization of motor torque and battery power. This improves the accuracy of battery state estimation and the adaptability of control strategies, optimizes overall machine energy efficiency, and extends battery life, making it suitable for intelligent control of pure electric bulldozers in construction machinery.

[0013] This disclosure discloses a method for state estimation and coordinated control of the power battery in a pure electric bulldozer. Within constraints, to achieve a high level of accuracy in describing complex nonlinear systems, a multi-objective coordinated optimization problem balancing battery life degradation and economic efficiency is constructed using model predictive control (MMC). The nonlinear optimization problem is transformed into a sequential quadratic programming (SQP) problem or a linearized QP problem. Through state-space discretization, a first-order Taylor expansion of the nonlinear model is performed at each sampling point to obtain a linear time-varying system. The objective function is then expanded into a standard quadratic form, ultimately transforming the MPC optimization problem into a quadratic programming problem for solution. The output power of the drive motor is used as the control variable. In the prediction time domain, the constructed nonlinear optimal control problem is transformed into a system of linear equations, which converge quickly using an iterative residual minimization method to improve computational efficiency. By integrating mechanism-driven and data-driven battery state estimation and whole-machine collaborative optimization control, the limitations of traditional control strategies that only consider instantaneous energy consumption are eliminated. The system accurately predicts the battery's state of equilibrium (SOH) and achieves multi-objective collaborative optimization control of the whole machine. This provides a complete, efficient, and forward-looking solution for addressing the economic, safety, and control algorithm adaptability issues faced by engineering machinery in practical applications. Attached Figure Description

[0014] The accompanying drawings, which form part of this disclosure, are used to provide a further understanding of this disclosure. The illustrative embodiments of this disclosure and their descriptions are used to explain this disclosure and do not constitute an undue limitation of this disclosure.

[0015] Figure 1 This is a schematic diagram of the topology of the whole machine multiphysics coupling model according to an embodiment of this disclosure; Figure 2This is a flowchart of the method for estimating the state of the power battery and coordinating control of a pure electric bulldozer according to an embodiment of this disclosure. Detailed Implementation

[0016] The present disclosure will be further described below with reference to the accompanying drawings and embodiments.

[0017] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of this disclosure. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this disclosure pertains.

[0018] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this disclosure. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms “comprising” and / or “including” are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0019] Example 1 One embodiment of this disclosure provides a method for estimating the state of the power battery and coordinating control of a pure electric bulldozer, the method steps of which include: Step 1: Obtain typical working condition data of the pure electric bulldozer, identify the typical working conditions, and construct a multi-physics coupling model of the whole machine based on the identification results. Step 2: Based on the multi-physics coupling model of the whole machine, considering vibration and temperature factors, construct the electro-thermal coupling model of the power battery; Step 3: Based on the power battery electro-thermal coupling model, analyze the key factors affecting the accuracy of battery SOH estimation, input the key factors into the gated recurrent unit network, and output the preliminary SOH estimation result; use the adaptive extended Kalman filter algorithm to correct the preliminary SOH estimation result, and obtain the final estimate of battery SOH. Step 4: Based on the final estimate of battery SOH, set the battery power boundary. Considering the impact of battery power boundary and battery SOH on battery life, establish a battery life degradation model. Construct the objective function with the minimum consumed battery power and equivalent battery life degradation cost. Use model predictive control method combined with generalized minimum residual numerical algorithm to solve quickly, and achieve coordinated optimization of motor torque and battery power.

[0020] As one embodiment, the battery state estimation and coordinated control method for a pure electric bulldozer disclosed herein comprehensively considers the SOP boundary and battery life degradation, constructs a multi-objective optimization control problem weighted by energy consumption and lifespan, and employs Model Predictive Control (MPC) combined with a generalized minimum residual numerical algorithm for rapid solution, thereby achieving coordinated optimization of motor torque and battery power. The specific implementation process of this disclosure is as follows: Step 1: Obtain typical working condition data of the pure electric bulldozer, identify the typical working conditions, and construct a multi-physics coupling model of the whole machine based on the identification results.

[0021] To accurately reflect the dynamic response of the entire machine and key components under complex working conditions, achieve accurate SOH estimation, and optimize energy management strategies, this disclosure requires the construction of a high-fidelity multi-physics coupling model of the pure electric bulldozer. The steps include: Step 11: Acquisition of typical operating condition data.

[0022] This disclosure presents a real-vehicle test conducted on a pure electric bulldozer prototype experimental platform. The test covered typical working conditions such as earthmoving, leveling, slope operation, and high-speed relocation. Data from these typical working conditions was collected at a set sampling frequency via CAN bus and sensor (IMU) modules. This data included vehicle motion status data, operational information, system status data, and environmental information. The typical working condition data was preprocessed and converted into a standard time-series dataset. The vehicle motion status data included longitudinal vehicle speed. v acceleration a and driving trajectory s The working information includes the pressure of the hydraulic cylinder for lifting the shovel. P 1. Tilting hydraulic cylinder pressure P 2 and blade displacement X System status data includes the output torque of the drive motor. T Rotation speed w Total current I ,Voltage U and temperature T 1. Environmental information includes ambient temperature. T 2 and platform tilt angle θ.

[0023] As one example, the sampling frequency can be set to 10Hz.

[0024] Furthermore, the preprocessing operations include moving median filtering for noise reduction, linear interpolation for point filling, and timestamp alignment, transforming the raw data into a standard time series dataset. D .

[0025] Step 12: Identify typical operating conditions and construct a multiphysics coupling model of the whole machine based on the identification results. The steps include: (1) Extract multi-dimensional features from the discrete load fragment set.

[0026] First, the time window method is used to process the dataset. D Segmentation is performed based on the steady-state criterion: standard deviation of vehicle speed. σ v <0.5 km / h Working criterion: Maximum hydraulic pressure of the shovel blade P max >(System rated pressure threshold), the dataset D Divided into n A set of discrete load segments S = { S 1, S 2, , S n}. In the set S Four dimensions were extracted, resulting in 16 multi-dimensional features related to optimal performance, including kinematic features, power system features, working device features, and energy features. Among these, the kinematic features include average vehicle speed. 、 Maximum speed v max average acceleration and the percentage of idling time P The characteristics of the power system include the average torque of the motor. Motor torque standard deviation σ T Average battery output power and battery peak power P b,max The working device features include hydraulic average pressure. Hydraulic peak pressure P h,max Work intensity index I = . v and the frequency of shovel movements f Energy characteristics include total energy consumption Current I and the proportion of regenerated energy recovered η .

[0027] This disclosure constructs a set of feature vectors for load segments based on the above 16-dimensional multi-dimensional features. X .

[0028] (2) Principal component analysis is performed based on sensitivity to obtain working condition characteristics.

[0029] Specifically, since different features have varying impacts on overall performance, directly analyzing all features can easily affect clustering results. Therefore, energy consumption is considered... E Battery peak power P b,max For output performance indicators Y ,make Y = f ( X The sensitivity analysis method (Sobol) is used to calculate the multi-dimensional features of the input. X j First-order response of the system S j With the total effect index S Tj :

[0030] in, V j For the first j The variance contribution of each feature acting alone; V jk For the first j , k The variance contribution of each feature interaction; p This represents the total number of features (16 initial features). Indicates except the first j The set of features other than the first feature.

[0031] reserve S Tj Features greater than 0.05 constitute a subset. , n For the number of payload segments, q This represents the number of features retained after filtering. X s Vector standardization Principal component analysis (PCA) is used to solve for the covariance matrix and perform eigenvalue decomposition.

[0032] in, λ j For the first j Each feature is worth contributing to the variance. v j This is the corresponding feature vector.

[0033] Will λ j Sort in descending order and select the smallest. m Make:

[0034] Constructing the principal component matrix This is used to obtain the PCA score matrix Z, which displays the operating condition characteristics of the driving system performance changes: .

[0035] (3) The working condition characteristics are clustered using K-means++. Through k-means clustering identification, working condition patterns with similar energy consumption and dynamic characteristics are obtained. Based on the clustering results, a database of typical scenario working conditions is constructed.

[0036] This disclosure aims to identify typical work scenarios and conditions by using the work condition feature Z as input features and employing K-means++ clustering to segment the work condition features. During the initialization phase, probability sampling is used to select... k An initial centroid is obtained by minimizing the sum of squares within the cluster. J :

[0037] in, The cluster mean vector, for n Each cluster is represented by a single data point. Using k-means clustering, in the sense of Euclidean distance, all segments are grouped into clusters. m The samples were divided into n Each cluster represents a working condition with similar energy consumption and dynamic characteristics. Finally, based on the clustering results, a database of typical scenario working conditions is constructed.

[0038] (4) Based on the database of typical scenario working conditions, analyze the working characteristics and dynamic response characteristics of the main powertrain and working device of the engineering machinery and construct a multi-physics coupling model of the whole machine.

[0039] Based on the established database of typical operating conditions, a multiphysics coupling model of the entire machine is built, and the topology of the entire machine is as follows: Figure 1 As shown, it includes a power battery pack, a motor controller, a drive motor, and drive wheels, which are connected by electrical and mechanical connections.

[0040] This study investigates the working characteristics and dynamic response of the main powertrain and operating devices of construction machinery. Kinematic and dynamic analyses of the power transmission system are performed under different operating modes, and a whole-machine dynamic model is built to simulate the dynamic behavior of the machine under typical operating conditions. Simultaneously, based on battery / motor test data, a multi-physics coupled model of mechanical, thermal, and electrochemical fields is constructed. For bulldozers using pure electric drive, a multi-physics coupled model of the motor of the motor (mechanical, electrical, and magnetic fields) is built to analyze the power output characteristics of the motor under different loads and speeds. Finally, based on relevant data from real-vehicle tests, the models of key components are verified and refined.

[0041] Ultimately, a multi-physics coupling model of a pure electric bulldozer suitable for overall machine control and optimization was built.

[0042] Step 2: Based on the multiphysics coupling model of the whole machine, and considering vibration and temperature factors, construct the electro-thermal coupling model of the power battery. Specifically, the steps include: (1) Collect battery operating data under different cycle counts, discharge rates, ambient temperatures, and vibration levels. Use the Pearson correlation coefficient method to extract characteristic parameters, evaluate the correlation between each characteristic and SOH degradation, and screen out the characteristic quantities related to SOH degradation: Based on the battery experimental platform, operational data of the battery was collected under different cycle numbers, discharge rates, ambient temperatures, and vibration levels. Combined with a multiphysics coupling model of the entire machine, the battery operating boundary under typical working conditions was calculated, and the influence of different factors on battery capacity decay and internal resistance growth was analyzed to study the degradation law of SOH (Sodium Hydrocarbons).

[0043] Based on this, the Pearson correlation coefficient method was used to extract characteristic parameters and assess the correlation between each characteristic and SOH decline:

[0044] in, xi For the characteristic quantity in the th i The value of the sample yi This corresponds to the SOH value. The sample mean. This represents the average total SOH.

[0045] Furthermore, by calculating the correlation coefficients between different characteristics and SOH... r Size was used to filter out features associated with SOH decline.

[0046] (2) Describe the voltage response characteristics of the battery under different charging and discharging conditions, create an equivalent thermal model, associate the electrical characteristics with the thermal characteristics, and build a dual-state lumped parameter equivalent thermal model; based on the relevant data of the actual battery measurement, use parameter identification to obtain the power battery electric-thermal coupling model adapted to pure electric bulldozer.

[0047] Specifically, the characteristics of the battery, such as capacity, DC resistance, open-circuit voltage and temperature, are analyzed. A second-order equivalent circuit model is built to describe the voltage response characteristics of the battery under different charging and discharging conditions. An equivalent thermal model is created to reflect the heat and temperature distribution generated by the battery during charging, discharging and high load processes.

[0048] To link electrical and thermal properties, a dual-state lumped parameter equivalent thermal model is constructed to describe the electrochemical energy conversion and heat transfer process of the battery and simulate its dynamic response under different temperature, load, and environmental conditions.

[0049] After constructing the electro-thermal coupling model, the model parameters are identified using the variable forgetting factor least squares (VFF-RLS) method based on the relevant data measured by the battery. By using parameter identification, an electro-thermal coupling model adapted to pure electric bulldozers is obtained.

[0050] Step 3: Based on the power battery electro-thermal coupling model, analyze the key factors affecting the accuracy of battery SOH estimation, input the key factors into the gated recurrent unit network, and output the preliminary SOH estimation result; use the adaptive extended Kalman filter algorithm to correct the preliminary SOH estimation result and obtain the final SOH estimate of the battery. Specifically, the steps include: (1) Based on the power battery electro-thermal coupling model, use the Pearson correlation coefficient method to extract the key factors for SOH estimation based on voltage, current, temperature, internal resistance, and constant current charging time data, input the extracted key factors into the gated recurrent unit GRU network, and perform preliminary SOH estimation; This disclosure analyzes the key factors affecting the accuracy of battery SOH estimation based on the battery electro-thermal coupling model, such as temperature fluctuations, current surges, mechanical vibrations, and changes in ambient temperature. It uses the Pearson correlation coefficient method to extract key factors for SOH estimation based on data such as voltage, current, temperature, internal resistance, and constant current charging time.

[0051] Specifically, based on the established electro-thermal coupling model of the power battery, the key factors after screening are set as follows: x k ,Will x k The input is fed into a gated recurrent unit (GRU) network for preliminary estimation of SOH. The calculation of the GRU during this process can be simplified as follows:

[0052] In the formula, zt To update the door status, rt To reset the door status, t In the candidate hidden state, ht This is the hidden state, which is the initial estimate of SOH by GRU; Wz To update the gate weights, Wr To reset the gate weight, Wh The weights of the candidate hidden states; bz To update the gate bias term, br To reset the door, bh The bias term for the candidate hidden state.

[0053] (2) The preliminary estimate of SOH is used as the observation input of the adaptive extended Kalman filter algorithm. Combined with the electro-thermal coupling mechanism model, the prediction of battery capacity and internal resistance change trend is updated. The filter gain is adjusted according to the residual between prediction and observation, and the preliminary estimate of GRU is corrected. Finally, the accurate estimate of battery SOH is achieved.

[0054] Specifically, the output of the GRU ht As the observation input for the Adaptive Extended Kalman Filter (AEKF) algorithm, combined with the electro-thermal coupling mechanism model, the prediction of battery capacity and internal resistance change trends is updated. The filter gain is adjusted according to the residual between the prediction and the observation, and the preliminary estimation results of GRU are corrected, ultimately achieving an accurate estimation of the battery's state of harm (SOH).

[0055] Finally, under different operating conditions, the accuracy of state estimation is evaluated based on real vehicle data. The influence of different hyperparameters (such as GRU network structure, learning rate, etc.) on the accuracy of SOH estimation is studied, and reasonable hyperparameters are selected to lay the foundation for subsequent whole-machine collaborative optimization.

[0056] Step 4: Based on the final estimate of battery SOH, set the battery power boundary. Considering the impact of the battery power boundary and battery SOH on battery life, establish a battery life degradation model. Construct the objective function with the minimum consumed battery power and equivalent battery life degradation cost. Use model predictive control methods combined with a generalized minimum residual numerical algorithm to quickly solve the problem, achieving coordinated optimization of motor torque and battery power. Specifically, the steps include: (1) Design the power state boundary of the power battery based on the health status of the power battery and the working environment temperature.

[0057] Specifically, the power battery state of power (SOP) boundary is designed based on the accurate estimation of the battery SOH and the power battery health status and operating environment temperature. The core of the battery state of power (SOP) boundary is the instantaneous and continuous power capability boundary of the power battery.

[0058] The instantaneous power boundary of the battery is influenced by the battery's current voltage, current, and internal resistance, and is expressed as follows:

[0059] In the formula, Ps max( k ) represents the instantaneous power boundary. V ( k )for k Voltage at any moment V min is the lower limit of the safe discharge voltage. I max( k ) represents the upper limit of the current. R ( SOH, T The equivalent resistance is calculated considering the effects of SOH and temperature.

[0060] Furthermore, the continuous power boundary is limited by battery temperature, expressed as:

[0061] In the formula, Q ( SOH To account for the usable capacity after SOH, T max The maximum allowable temperature of the battery, T( k )for k Constantly monitor battery temperature. R t This is the thermal resistance.

[0062] Ultimately, the SOP (Start of Production) boundary for power batteries P ( k It is composed of instantaneous power and continuous power, that is:

[0063] The SOP constraint changes with the SOH and temperature. When performing whole-machine collaborative optimization, it is introduced as a hard constraint into the subsequent optimization problem as an MPC constraint to ensure that the motor power distribution does not exceed the battery's allowable range during the optimization process.

[0064] (2) Construct an MPC multi-objective collaborative optimization model; After designing the SOP boundary, this disclosure analyzes the impact of the SOP boundary and SOH on battery life. Based on this, a battery life degradation model is established, which is divided into cycle degradation and calendar degradation. The key parameters of the model are identified through actual measurement data.

[0065] This paper explains the impact mechanism between battery life degradation and overall machine control performance. Based on the typical working conditions of a bulldozer, simply pursuing low energy consumption may lead to an increase in the frequency of deep discharge, causing rapid degradation of State of Harm (SOH). On the other hand, adopting an overly conservative strategy may reduce work efficiency and economy. Therefore, in the stage of overall machine optimization control, this disclosure constructs an objective function by weighting the consumed battery power and the equivalent battery life degradation cost.

[0066] In the formula, Pb ( k )for k Battery output power at any given time, Δ SOH ( k (This is a prediction based on a battery life degradation model) k The amount of SOH degradation at time t, PThis is the conversion factor. Q and R As a weight matrix, the SOH degradation is expressed using a reduction factor. P This translates into an equivalent lifetime decay cost.

[0067] Furthermore, the constraints for each physical component are determined as follows: 1) SOP boundary constraints:

[0068] 2) Motor torque constraint:

[0069] 3) Current constraint:

[0070] Avoid overheating or damage to the motor due to overcurrent.

[0071] 4) Battery SOC range constraints:

[0072] Avoid overcharging and over-discharging.

[0073] 5) Temperature constraint:

[0074] Prevent performance degradation caused by high or low temperatures.

[0075] Within the aforementioned constraints, to achieve a high level of accuracy in describing complex nonlinear systems, a multi-objective collaborative optimization problem balancing battery life degradation and economic efficiency is constructed using model predictive control (MRC). The nonlinear optimization problem is transformed into a sequential quadratic programming (SQP) problem or a linearized QP problem. Through state-space discretization, a first-order Taylor expansion of the nonlinear model is performed at each sampling point to obtain a linear time-varying system. The objective function is then expanded into a standard quadratic form, ultimately transforming the MPC optimization problem into a quadratic programming problem for solution. The output power of the drive motor is used as the control variable. A fast iterative solution is achieved using the generalized minimum residual numerical algorithm (C / GMRES), i.e.:

[0076] In the formula, F ( x , u ) represents the nonlinear state residual constructed from state variables and control variables.

[0077] Within the prediction time domain, the constructed nonlinear optimal control problem is transformed into a system of linear equations, which converge quickly using an iterative residual minimization method to improve computational efficiency. During the process, speed and torque data collected from actual vehicles are used to identify and correct model parameters online, enabling the prediction model to dynamically adapt to changes in SOH (State of Hypoxia). Finally, through optimization, the optimal drive torque of the motor and battery power are obtained.

[0078] Finally, under different operating conditions and loads, the control parameters (such as weightings) of the objective function are adjusted. P Sensitivity analysis is performed by perturbing the parameters within a given interval and recording the instantaneous energy consumption. E Changes in SOH decay rate ΔSOH, etc., and calculate sensitivity indicators:

[0079] By dynamically adjusting parameters with high sensitivity, the robustness of the overall machine control strategy under different operating conditions and loads can be ensured.

[0080] Example 2 One embodiment of this disclosure provides a power battery state estimation and cooperative control system for a pure electric bulldozer, including: The whole vehicle model building module is used to acquire typical working condition data of the pure electric bulldozer, identify the typical working conditions, and build a multi-physics coupling model of the whole machine based on the identification results. The electrothermal coupling model construction module is used to construct an electrothermal coupling model of the power battery based on the multi-physics coupling model of the whole machine, taking into account vibration and temperature factors. The battery estimation module is used to analyze the key factors affecting the accuracy of battery SOH estimation based on the power battery electro-thermal coupling model. The key factors are input into the gated recurrent unit network and the output is a preliminary SOH estimation result. The preliminary SOH estimation result is corrected by an adaptive extended Kalman filter algorithm to obtain the final estimate of battery SOH. The collaborative optimization module is used to set the battery power boundary based on the final estimate of the battery SOH. It considers the impact of the battery power boundary and the battery SOH on the battery life, establishes a battery life degradation model, constructs the objective function with the minimum consumed battery power and the equivalent battery life degradation cost, and uses model predictive control method combined with generalized minimum residual numerical algorithm to solve it quickly, so as to achieve collaborative optimization of motor torque and battery power.

[0081] Example 3 One embodiment of this disclosure provides a computer program product, including a computer program that, when executed by a processor, implements the method for estimating the state of the power battery and coordinating control of a pure electric bulldozer.

[0082] Example 4 One embodiment of this disclosure provides a non-transitory computer-readable storage medium for storing computer instructions. When the computer instructions are executed by a processor, they implement the pure electric bulldozer power battery state estimation and cooperative control method.

[0083] Example 5 One embodiment of this disclosure provides an electronic device, including a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to implement the pure electric drive bulldozer power battery state estimation and cooperative control method.

[0084] This disclosure is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this disclosure. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0085] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0086] While the specific embodiments of this disclosure have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of this disclosure. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of this disclosure are still within the scope of protection of this disclosure.

Claims

1. A method for state estimation and coordinated control of the power battery of a pure electric bulldozer, characterized in that, include: Obtain typical working condition data of the pure electric bulldozer, identify the typical working conditions, and construct a multi-physics coupling model of the whole machine based on the identification results; Based on the multi-physics coupling model of the whole machine, and considering vibration and temperature factors, an electro-thermal coupling model of the power battery is constructed. Based on the electric-thermal coupling model of the power battery, the key factors affecting the accuracy of battery SOH estimation are analyzed. The key factors are input into the gated recurrent unit network and the preliminary SOH estimation results are output. The initial SOH estimate was corrected using an adaptive extended Kalman filter algorithm to obtain the final SOH estimate of the battery. Based on the final estimate of battery SOH, the battery power boundary is set. Considering the influence of the battery power boundary and battery SOH on battery life, a battery life degradation model is established. The objective function is constructed with the minimum consumed battery power and the equivalent battery life degradation cost. The model predictive control method is adopted and combined with the generalized minimum residual numerical algorithm to solve the problem quickly, so as to achieve the coordinated optimization of motor torque and battery power.

2. The method for state estimation and cooperative control of the power battery of a pure electric bulldozer as described in claim 1, characterized in that, Based on the various typical working conditions of pure electric bulldozers, such as earthmoving, fine leveling, slope operation, and high-speed transfer, typical working condition data are collected through the CAN bus and sensor modules at a set sampling frequency. This includes vehicle motion status data, working information, system status data, and environmental information. The typical working condition data is then preprocessed and converted into a standard time series dataset.

3. The method for state estimation and coordinated control of the power battery of a pure electric bulldozer as described in claim 1, characterized in that, The process of identifying typical operating conditions and constructing a multiphysics coupling model of the entire machine based on the identification results includes: Multi-dimensional features are extracted from a discrete set of load segments, and principal component analysis is performed based on sensitivity to obtain the working condition features. The operating condition characteristics are identified using the K-means++ clustering method. Through k-means clustering, operating condition patterns with similar energy consumption and dynamic characteristics are obtained. Based on the clustering results, a database of typical scenario operating conditions is constructed. Based on a database of typical operating conditions, the working characteristics and dynamic response characteristics of the main powertrain and working devices of the engineering machinery are analyzed, and a multi-physics coupling model of the whole machine is constructed.

4. The method for estimating the state of the power battery and coordinating control of a pure electric bulldozer as described in claim 1, characterized in that, The aforementioned electro-thermal coupling model of the power battery, based on the multi-physics coupling model of the whole machine and considering vibration and temperature factors, includes: Battery operation data were collected under different cycle numbers, discharge rates, ambient temperatures and vibration levels. The Pearson correlation coefficient method was used to extract characteristic parameters, evaluate the correlation between each characteristic and SOH degradation, and screen out the characteristic quantities related to SOH degradation. Describe the voltage response characteristics of the battery under different charge and discharge conditions, create an equivalent thermal model, associate the electrical characteristics with the thermal characteristics, and build a two-state lumped parameter equivalent thermal model. Based on relevant data from actual battery measurements, parameter identification was used to obtain an electro-thermal coupling model of the power battery adapted to pure electric bulldozers.

5. The method for state estimation and coordinated control of the power battery of a pure electric bulldozer as described in claim 1, characterized in that, Based on the electro-thermal coupling model of the power battery, the Pearson correlation coefficient method is used to extract key factors for SOH estimation based on voltage, current, temperature, internal resistance, and constant current charging time data. The extracted key factors are then input into the gated recurrent unit (GRU) network for preliminary SOH estimation. The preliminary estimate of SOH is used as the observation input of the adaptive extended Kalman filter algorithm. Combined with the electro-thermal coupling mechanism model, the prediction of battery capacity and internal resistance change trends is updated. The filter gain is adjusted according to the residual between the prediction and observation to correct the preliminary estimate of GRU, and finally the accurate estimate of battery SOH is achieved.

6. The method for state estimation and cooperative control of the power battery of a pure electric bulldozer as described in claim 1, characterized in that, Based on the health status of the power battery and the operating temperature, the power battery power state boundary is designed. The core of the boundary is the instantaneous and continuous power capability boundary of the power battery. The power battery SOP boundary is composed of the instantaneous power and continuous power, which is used as one of the constraints of multi-objective optimization. A battery life degradation model is constructed. In the stage of whole machine optimization control, the objective function is constructed by weighted summing of the consumed battery power and the equivalent battery life degradation cost. Through state space discretization, a first-order Taylor expansion is performed on the nonlinear model at each sampling point to obtain a linear time-varying system. Then, the objective function is expanded into a standard quadratic form, and the MPC optimization problem is finally transformed into a quadratic programming problem to be solved.

7. A power battery state estimation and cooperative control system for a pure electric bulldozer, characterized in that, include: The whole vehicle model building module is used to acquire typical working condition data of the pure electric bulldozer, identify the typical working conditions, and build a multi-physics coupling model of the whole machine based on the identification results. The electrothermal coupling model construction module is used to construct an electrothermal coupling model of the power battery based on the multi-physics coupling model of the whole machine, taking into account vibration and temperature factors. The battery estimation module is used to analyze the key factors affecting the accuracy of battery SOH estimation based on the power battery electro-thermal coupling model, input the key factors into the gated recurrent unit network, and output the preliminary SOH estimation results. The initial SOH estimate was corrected using an adaptive extended Kalman filter algorithm to obtain the final SOH estimate of the battery. The collaborative optimization module is used to set the battery power boundary based on the final estimate of the battery SOH. It considers the impact of the battery power boundary and the battery SOH on the battery life, establishes a battery life degradation model, constructs the objective function with the minimum consumed battery power and the equivalent battery life degradation cost, and uses model predictive control method combined with generalized minimum residual numerical algorithm to solve it quickly, so as to achieve collaborative optimization of motor torque and battery power.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the method for estimating the state of the power battery and coordinating control of a pure electric bulldozer as described in any one of claims 1-6.

9. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium is used to store computer instructions, which, when executed by a processor, implement the pure electric drive bulldozer power battery state estimation and cooperative control method as described in any one of claims 1-6.

10. An electronic device, characterized in that, include: The device includes a processor, a memory, and a computer program; wherein the processor is connected to the memory, and the computer program is stored in the memory. When the electronic device is running, the processor executes the computer program stored in the memory to enable the electronic device to perform the method for estimating and coordinating the state of the power battery of a pure electric bulldozer as described in any one of claims 1-6.

Images