Data and working condition based control system and method for kinetic energy recovery of new energy micro truck

Abstract

The application discloses a new energy micro truck kinetic energy recovery control system and method based on data and working conditions. The application makes full use of long-term vehicle operation data, generates differentiated strategies for typical routes, loads and driving styles, significantly improves the energy recovery efficiency under typical working conditions, and has the technical effect of iterative optimization over time. The application structures the complex operation scene through the working condition label, which is convenient for the cloud to respectively count, optimize the upper limit of the recovery torque, the exit threshold, the slope / attachment compensation and other parameters under each type of working condition, realizes fine and adaptive control for the complex working conditions of the micro truck, adopts the candidate strategy generation and user multi-strategy selection mechanism of vehicle-cloud cooperation, explicitly introduces safety constraints in the execution phase, ensures the braking safety and stability under the conditions of strategy iteration and cloud unavailability, and reduces the calibration and maintenance cost, improves the scalability and engineering landing of the system.

Description

Technical Field

[0001] This invention relates to a kinetic energy recovery control system and method for new energy micro-trucks. Background Technology

[0002] New energy micro-trucks need to operate under typical conditions such as urban distribution, suburban delivery, and rural feeder routes. These conditions are characterized by large variations in load, frequent start-stop operations, numerous slopes, and complex road conditions. The energy recovery control schemes commonly used in new energy vehicles generally fall into two categories:

[0003] The first energy recovery control scheme allocates the proportion of recovered torque to hydraulic braking torque within the target torque based on the hazard level and a preset energy recovery control table. Generally, the higher the hazard level, the smaller the recovery proportion, while the energy recovery control table is fixed in calibration. This scheme relies heavily on a preset control table or fixed rules, employing a static calibration strategy, making it difficult to adapt to different routes, loads, and driver styles.

[0004] The second energy recovery control scheme uses information such as vehicle speed, vehicle weight, state of charge (SOC), battery temperature, road surface smoothness, gradient, driving conditions, and hazardous conditions to calculate the "most reasonable" energy recovery intensity in real time, and executes it through recovery in the early stage of pedal travel and mechanical braking in the later stage. Compared with the first energy recovery control scheme, although this scheme utilizes information such as current vehicle speed, SOC, temperature, gradient, and hazard level, it does not systematically utilize long-term historical operating data for parameter model construction and iteration. Summary of the Invention

[0005] The technical problem this invention aims to solve is that existing kinetic energy recovery control schemes lack a working condition label library and parameter mapping mechanism for delivery conditions involving multiple loads, routes, and slopes for micro-trucks, and it is difficult for users to choose a suitable strategy based on their own preferences (energy saving priority, smoothness priority, etc.).

[0006] To address the aforementioned technical problems, the first aspect of the present invention discloses a new energy micro-truck kinetic energy recovery control system based on data and operating conditions, characterized in that it employs a vehicle-side... cloud The user interface has a three-layer architecture, including a vehicle-side kinetic energy recovery control device, a cloud-based strategy generation device, and a user interaction and strategy selection device, wherein: The vehicle-side kinetic energy recovery control device further includes: The data acquisition unit is used to collect cruise data of the target new energy micro-truck; The condition identification unit is used to calculate the load level label L_load, gradient level label L_grade, adhesion level label L_mu, driving state type label L_state, driving style level label L_style, and SOC interval label L_SOC based on the cruise data collected by the data acquisition unit, and generate the condition label vector L = {L_load, L_grade, L_mu, L_state, L_style, L_SOC, route_id}. The communication unit is used for bidirectional transmission of data and policy parameters with the cloud-based policy generation device. The policy storage unit is used to locally cache a subset of the policy parameter model library, the policy parameter set S_active, and the basic security policy S_safe_base issued from the cloud policy generation device; The execution control unit is used to calculate the expected total deceleration torque T_req, motor recovery torque T_rec_cmd, and hydraulic braking torque T_hyd_cmd based on the driver's braking requirements and strategy parameter set S_active of the target new energy micro-truck, and send the motor recovery torque T_rec_cmd to the vehicle controller or motor controller, and send the hydraulic braking torque T_hyd_cmd to the braking system. The safety management unit is used to monitor in real time the estimated value of road surface adhesion coefficient φ_est, battery temperature T_bat, battery receiveable power P_chg_max, braking system fault flags and communication status. Under the condition that the safety constraints are not met, it performs online derating or rollback to the basic safety policy S_safe_base for some parameters in the strategy parameter set S_active. The cloud-based policy generation device further includes: The data receiving and storage unit is used to receive historical operation data packets from multiple new energy micro-trucks. The historical operation data packets contain all the cruise data of each new energy micro-truck collected by the data acquisition unit. The working condition clustering and label generation unit is used to extract features and cluster historical sample data received by the data receiving and storage unit, generate a discrete set of working condition labels, and maintain label definitions and thresholds. The strategy parameter modeling unit is used to statistically analyze the performance indicators of candidate strategy parameters under each combination of working condition labels generated by the working condition clustering and label generation unit, and to obtain strategy parameter sets of various styles using optimization algorithms. The model library management unit is used to maintain the mapping relationship between each type of working condition label combination and multiple styles of strategy parameter sets, and to build and obtain the strategy parameter model library. The strategy parameter matching and distribution unit is used to respond to the request of the vehicle-side kinetic energy recovery control device of the target new energy micro-truck. Based on the current working condition label vector L_cur and preference information, it performs working condition indexing and matching from the strategy parameter model library, selects or generates several sets of candidate strategies as subsets of the strategy parameter model library, packages them and distributes them to the vehicle-side kinetic energy recovery control device of the target new energy micro-truck. The user interaction and strategy selection device is used to display the candidate strategy summary in the strategy parameter model library subset, receive user / fleet strategy selection, select a set of candidate strategies from the strategy parameter model library subset based on the user / fleet strategy selection as the strategy parameter set S_active, and deliver the strategy parameter set S_active to the vehicle-side kinetic energy recovery control device.

[0007] Preferably, the cruise data includes vehicle speed v, longitudinal deceleration a, wheel speed ω, gear / drive mode m, road gradient θ, ambient temperature T_env, road surface adhesion coefficient estimate φ_est, vehicle mass or load estimate M_est, battery state of charge (SOC), battery temperature T_bat, battery receiveable power P_chg_max, current motor regenerative torque T_rec_cur, brake pedal travel s_brk, braking duration t_brk, braking frequency f_brk, whether emergency braking is required (flag_emg), route identifier (route_id), mileage location d, driver identifier (driver_id), and timestamp (time).

[0008] Preferably, based on the estimated vehicle mass or load value M_est and the preset curb weight M_0, the load ratio = (M_est - M_0) / M_0 is calculated, and according to the interval of the calculated load ratio, the current load level is discretized into at least one of light load, medium load and heavy load to obtain the load level label L_load. Based on the average value of the collected road slope θ within a preset time window, the current slope is divided into at least one of the following: flat road, uphill or downhill, and optional long downhill, to obtain the slope grade label L_grade. Based on the collected road surface adhesion coefficient estimate φ_est, the current adhesion level is divided into at least one of high adhesion, medium adhesion and low adhesion according to the preset adhesion coefficient range, so as to obtain the adhesion level label L_mu. Within a preset time or mileage window, the vehicle speed v, longitudinal deceleration a, braking duration t_brk, and braking frequency f_brk are statistically analyzed. Based on the average vehicle speed, the degree of vehicle speed fluctuation, and the braking frequency per unit mileage, the current driving state is divided into at least one of frequent start-stop, long downhill, cruise, or other mixed conditions to obtain the driving state type label L_state. Based on the longitudinal acceleration and deceleration characteristics, accelerator pedal usage characteristics, and braking operation characteristics of the driver identification within a preset mileage range, a driving style score is calculated, and the current driving style is divided into at least one of economic / mild, normal, or aggressive according to the interval in which the driving style score is located, so as to obtain the driving style level label L_style. Based on the battery state of charge (SOC), battery temperature (T_bat), and battery receiveable power (P_chg_max), the current SOC state is divided into at least one of a low SOC range, a medium SOC range, or a high SOC range to obtain the SOC range label L_SOC.

[0009] The second aspect of the technical solution of this invention discloses a data- and operating condition-based kinetic energy recovery control method for new energy micro-trucks. The method employs the aforementioned new energy micro-truck kinetic energy recovery control system, characterized in that it includes a vehicle-side portion operating in the vehicle-side kinetic energy recovery control device and a cloud-side portion operating in the cloud-based strategy generation device, wherein: The vehicle-end portion includes the following steps: Step 1: Collect cruise data of new energy micro-trucks, package the collected cruise data and upload it to the cloud strategy generation device; Step 2: Within each control cycle, the operating condition identification unit forms the current operating condition label vector L_cur={L_load,L_grade,L_mu,L_state, L_style,L_SOC,route_id} based on the real-time cruise data collected by the data acquisition unit. Then, it packages the previous operating condition label vector L_cur together with the vehicle identifier vehicle_id, route identifier route_id, driver identifier driver_id, and user preset preferences into a request message Req and sends it to the cloud policy generation device. Step 3: The vehicle-side kinetic energy recovery control device receives N sets of candidate strategy parameters S_k, where N≥2, k=1…N, fed back by the cloud-based strategy generation device; Step 4: The user selects one set of candidate policy parameters from N sets of candidate policy parameter sets S_k through the user interaction and policy selection device as the policy parameter set S_active; Step 5: The execution control unit loads the strategy parameter set S_active, calculates the expected total deceleration torque T_req, motor recovery torque T_rec_cmd, and hydraulic braking torque T_hyd_cmd, and sends the motor recovery torque T_rec_cmd to the vehicle controller or motor controller, and sends the hydraulic braking torque T_hyd_cmd to the braking system. Step 6: Monitor the estimated road surface adhesion coefficient φ_est, battery temperature T_bat, battery receiveable power P_chg_max, braking system fault flags, and communication status in real time. If safety constraints are not met, derate some parameters in the strategy parameter set S_active online or roll back to the basic safety strategy S_safe_base. After returning to normal, restore to the strategy parameter set S_active by annealing or request a new strategy. The cloud component includes the following steps: Step A: The cloud-based strategy generation device collects historical operation data packets from multiple new energy micro-trucks to construct operation task samples; Step B: Based on the extracted feature vector X={average load ratio λ_load, gradient statistics θ_mean / θ_max, speed distribution parameters, start-stop frequency f_start, road surface adhesion statistics φ_est, SOC distribution, braking frequency}, for the same route or driver, the running task samples are divided into K typical working conditions, and the working condition label vector L={L_load,L_grade,L_mu,L_state,L_style,L_SOC,route_id} is obtained for each typical working condition. Step C: For each combination of working conditions, use multi-objective optimization or heuristic search to find several Pareto optimal solutions for all historical strategy configurations according to budget constraints, and cluster these Pareto optimal solutions according to preferences to form different styles of strategy parameter sets, and then store them in the strategy parameter model library. Step D: After receiving the request message Req, the cloud-based strategy generation device matches the current operating condition label vector L_cur with the operating condition label vector in the strategy parameter model library, and matches the preference information with different styles in the strategy parameter model library. The obtained matched strategy parameter set is then sent to the vehicle-side kinetic energy recovery control device as a subset of the strategy parameter model library.

[0010] Preferably, in step 1, when collecting cruise data of the new energy micro-truck, if a braking event or a significant deceleration event is detected, the start and end times of the corresponding cruise data are additionally recorded, and the current set of cruise data is marked as a "critical deceleration segment". Only the cruise data marked as "critical deceleration segment" is packaged and uploaded. A significant deceleration event refers to an event in which the longitudinal deceleration a < speed threshold a_th.

[0011] Preferably, in step 1, the packaging and uploading of cruise data is triggered only if any of the following conditions are met: Condition 1) The current single delivery task of the new energy micro-truck has ended; Condition 2) Cumulative operating mileage ΔD ≥ mileage threshold D_th; Condition 3) The change in SOC |ΔSOC| ≥ the SOC threshold SOC_th; Condition 4) The amount of locally cached data exceeds the capacity threshold.

[0012] Preferably, in step 3, each set of candidate strategy parameter sets S_k includes: Vehicle speed segment {v_i} and corresponding maximum regenerative torque limit T_rec_max(i); The rate of change of regenerative torque with respect to pedal travel / deceleration, k_rec(i); Low-speed exit threshold v_cut_low, SOC exit threshold SOC_cut; The slope compensation coefficient calculation function is K_grade(θ), and the low adhesion reduction coefficient calculation function is K_mu(φ_est). Mechanical braking compensation ratio α_hyd(i) and switching point s_hyd(i); And, strategic style identifiers.

[0013] Preferably, in step 3, the vehicle-side kinetic energy recovery control device also receives brief indicators; In step 4, if the user does not make a selection, a candidate strategy parameter set is automatically selected from the N candidate strategy parameter sets S_k according to the pre-set priority based on the brief indicators.

[0014] Preferably, in step 5, the desired total deceleration torque T_req is calculated using the following formula: T_req = r_w m_est a_req Where r_w is the effective radius of the wheel, m_est is the estimated mass or load of the vehicle, and a_req is the expected deceleration; The motor recovery torque T_rec_cmd is calculated using the following formula: T_rec_cmd=min(T_req·k_rec(i),T_rec_max_corr,T_rec_bat(P_chg_max)); In the formula: T_rec_max_corr is the upper limit of the maximum recovery torque under the current operating conditions, T_rec_max_corr = T_rec_max(i) K_grade(θ') K_mu(φ_est') substitutes the real-time measured road slope θ' into the slope compensation coefficient calculation function K_grade(θ) to obtain the current slope compensation coefficient K_grade(θ'). K_mu(φ_est') substitutes the real-time estimated road surface adhesion coefficient φ_est' into the low adhesion depreciation coefficient calculation function K_mu(φ_est) to obtain the current adhesion compensation coefficient K_mu(φ_est'). T_rec_bat(P_chg_max) is the maximum regenerative torque allowed under the current operating conditions. It is obtained by the execution control unit through table lookup or interpolation calculation in a pre-calibrated "battery charging capacity - regenerative torque" mapping table based on the current battery state of charge (SOC), battery temperature (T_bat), battery receptive power (P_chg_max), and current vehicle speed (v). The "battery charging capacity - regenerative torque" mapping table is obtained through offline testing and simulation calibration. The hydraulic braking torque T_hyd_cmd is calculated using the following formula: T_hyd_cmd=T_req T_rec_cmd.

[0015] Preferably, in step 6, the conditions that do not meet the security constraints include: The road surface adhesion coefficient φ_est is lower than the adhesion threshold φ_low; Or, the battery temperature T_bat exceeds the preset safety range; Alternatively, the current battery's receiveable power P_chg_max is less than the preset power lower limit P_min; Alternatively, a fault indicator may appear in the braking system / critical sensors; Alternatively, the communication unit may be unable to connect to the cloud for an extended period of time.

[0016] Compared with existing kinetic energy recovery control schemes, the present invention has the following advantages: (1) In view of the problem that existing kinetic energy recovery control schemes only calculate the kinetic energy recovery intensity in real time at the vehicle end based on the current danger level or current perception information, and lack a mechanism for statistical modeling and continuous optimization of the strategy using long-term historical operating data, this invention constructs a kinetic energy recovery strategy model library based on multi-vehicle historical operating data and cloud big data analysis, and achieves the technical effect of making full use of long-term vehicle operating data, generating differentiated strategies for typical routes, loads and driving styles, significantly improving energy recovery efficiency under typical working conditions, and being able to iteratively optimize over time; (2) Existing kinetic energy recovery control schemes only make instantaneous judgments on operating conditions (such as hazard level, whether frequent start-stop, etc.), and do not form reusable "operating condition tags". The mapping of "strategy parameters" makes it difficult for strategies to be finely matched to delivery scenarios with multiple loads, slopes, and routes for micro-trucks. This invention constructs a mapping relationship between a micro-truck-specific working condition label system (multi-dimensional labels such as load / slope / adhesion / driving status / driving style / route, etc.) and the strategy parameter set. By structuring complex operating scenarios through working condition labels, it is easy for the cloud to separately count and optimize parameters such as the upper limit of recovery torque, exit threshold, and slope / adhesion compensation under each type of working condition, so as to achieve fine-grained and adaptive control for complex working conditions of micro-trucks. (3) To address the problem that existing kinetic energy recovery control schemes typically rely on the controller to automatically provide a single kinetic energy recovery intensity, preventing users from selecting strategies based on their energy-saving / smoothness / safety preferences, thus affecting availability and fleet management flexibility, this invention adopts a vehicle-to-everything (V2X) approach. The cloud-based collaborative candidate strategy generation and user multi-strategy selection mechanism generates multiple styles of strategy parameter sets (energy saving priority, smoothness priority, safety priority, etc.) for the same working condition tag combination in the cloud. The vehicle interface displays summary indicators for drivers or fleets to select strategies, improving the interpretability and user experience of the system. (4) To address the problem that existing kinetic energy recovery control schemes mainly output the recovery intensity directly at the vehicle end according to a fixed table or real-time algorithm, lacking a unified safety constraint and degradation mechanism after being combined with the cloud parameter model, and the unclear safety boundary when the adhesion is low, the battery is limited, or the controller is abnormal, the present invention provides vehicle-end recovery based on the issued strategy parameter set. Hydraulic collaborative control and online correction / degradation of safety constraints explicitly introduce safety constraints such as adhesion conditions, battery acceptable power, and system faults during the execution phase. When necessary, the recovery torque is reduced online, and the basic safety strategy is exited or rolled back in advance to ensure braking safety and stability even when the strategy is iterated and the cloud is unavailable. (5) In view of the fact that most existing kinetic energy recovery control schemes are one-time calibration or single-vehicle online algorithms, lacking a unified learning and model sharing mechanism across vehicles and routes, resulting in high calibration costs and difficulty in transferring strategies between fleets, this invention adopts a vehicle-based approach for micro-truck delivery scenarios. Cloud data closed-loop architecture (collection) Report Modeling Issued implement (Further data collection) establishes a closed loop from vehicle-side data collection and cloud-based modeling to policy distribution, execution, and feedback, enabling optimization experience from a particular route / driver to be transferred to the entire fleet, reducing calibration and maintenance costs, and improving system scalability and engineering feasibility. Attached Figure Description

[0017] Figure 1 This is a flowchart of a new energy micro-truck kinetic energy recovery control method based on data and operating conditions, as disclosed in an embodiment of the present invention. Detailed Implementation

[0018] The present invention will be further illustrated below with reference to specific embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. Furthermore, it should be understood that after reading the teachings of this invention, those skilled in the art can make various alterations or modifications to the invention, and these equivalent forms also fall within the scope defined by the appended claims.

[0019] The first aspect of this invention discloses a new energy micro-truck kinetic energy recovery control system based on data and operating conditions, employing a vehicle-side... cloud The user interface has a three-layer architecture, including a vehicle-side kinetic energy recovery control device, a cloud-based strategy generation device, and a user interaction and strategy selection device.

[0020] 1) Vehicle-side kinetic energy recovery control device 1) Data acquisition unit: Collect vehicle speed v (km / h) and longitudinal deceleration a (m / s²) 2 The parameters include: wheel speed ω, gear / drive mode m, road slope θ (° or %), ambient temperature T_env (°C), road surface adhesion coefficient estimate φ_est (dimensionless), vehicle mass or load estimate M_est (kg), battery state of charge SOC (%), battery temperature T_bat (°C), battery receiveable power P_chg_max (kW), current motor recovery torque T_rec_cur (N·m), brake pedal travel s_brk (%), braking duration t_brk (s), braking frequency f_brk (times / km), emergency braking flag_emg (Boolean), route identifier route_id, mileage location d (km), driver identifier driver_id, timestamp time, etc.

[0021] 2) Operating condition identification unit: Based on the real-time data collected by the data acquisition unit, the current load level label L_load, gradient level label L_grade, adhesion level label L_mu, driving state type label L_state (frequent starts / stops / long downhill / cruise), driving style level label L_style, and SOC range label L_SOC are calculated, generating a working condition label vector L = {L_load, L_grade, L_mu, L_state, L_style, L_SOC, route_id}, where: Based on the estimated vehicle mass or load value M_est and the preset curb weight M_0, calculate the load ratio = (M_est-M_0) / M_0, and according to the interval of the calculated load ratio, discretize the current load level into at least one of light load, medium load and heavy load to obtain the load level label L_load. Based on the average value of the collected road slope θ within a preset time window, the current slope is divided into at least one of the following: flat road, uphill or downhill, and optional long downhill, to obtain the slope grade label L_grade. Based on the collected road surface adhesion coefficient estimate φ_est, the current adhesion level is divided into at least one of high adhesion, medium adhesion and low adhesion according to the preset adhesion coefficient range, so as to obtain the adhesion level label L_mu. Within a preset time or mileage window, the vehicle speed v, longitudinal deceleration a, braking duration t_brk, and braking frequency f_brk are statistically analyzed. Based on the average vehicle speed, the degree of vehicle speed fluctuation, and the braking frequency per unit mileage, the current driving state is classified into at least one of frequent start-stop, long downhill, cruise, or other mixed conditions to obtain the driving state type label L_state. Based on the longitudinal acceleration and deceleration characteristics, accelerator pedal usage characteristics, and braking operation characteristics of the driver identification within a preset mileage range, a driving style score is calculated, and the current driving style is divided into at least one of economic / mild, normal, or aggressive according to the interval in which the driving style score is located, so as to obtain the driving style level label L_style. Based on the battery state of charge (SOC), battery temperature (T_bat), and battery power (P_chg_max), the current SOC state is divided into at least one of a low SOC range, a medium SOC range, or a high SOC range to obtain the SOC range label L_SOC. route_id is the route identifier, obtained by encoding the origin and destination addresses using BASE_64; The generated operating condition label vector L is output to the strategy parameter matching and distribution unit of the cloud-based strategy generation device.

[0022] 3) Communication unit: It enables bidirectional transmission of data and policy parameters with the cloud-based policy generation device via cellular or other vehicle-to-everything (V2X) communication modules, supporting encryption, retransmission of interrupted data, and version number management.

[0023] 4) Policy storage unit: The local cache contains a subset of the strategy parameter model library issued by the cloud-based strategy generation device and the strategy parameter set S_active fed back by the user interaction and strategy selection device. At the same time, it stores a set of built-in basic safety policies S_safe_base. The basic safety policy S_safe_base is a set of kinetic energy recovery control parameters that are pre-generated and solidified under offline conditions based on the vehicle and component design parameters, braking regulations and enterprise safety specifications, simulation calculation results under typical and extreme working conditions, and actual test calibration data.

[0024] 5) Execution control unit: Receive the driver's braking request (e.g., pedal travel / ACC deceleration request) and the strategy parameter set S_active, and calculate the expected total deceleration torque T_req, motor regenerative torque T_rec_cmd, and hydraulic braking torque T_hyd_cmd; The motor recovery torque T_rec_cmd is sent to the vehicle controller or motor controller, and the hydraulic braking torque T_hyd_cmd is sent to the braking system (e.g., ESP module).

[0025] 6) Safety Management Unit: Real-time monitoring of road surface adhesion coefficient estimate φ_est, battery temperature T_bat, battery receiveable power P_chg_max, braking system fault signs, communication status, etc. When security constraints are not met, some parameters in the policy parameter set S_active are derated online or rolled back to the basic security policy S_safe_base.

[0026] (ii) Cloud-based strategy generation device 1) Data receiving and storage unit: The system receives historical operation data packets from multiple new energy micro-trucks. These packets contain all data collected by the data acquisition unit of the aforementioned vehicle-side kinetic energy recovery control device and are stored as historical sample data in a distributed database indexed by vehicle, route, time, etc.

[0027] 2) Working condition clustering and label generation unit: Feature extraction and clustering are performed on historical sample data in a distributed database to generate a discrete set of chemical condition labels, and label definitions and thresholds are maintained.

[0028] 3) Strategy parameter modeling unit: For each combination of working condition labels generated by the working condition clustering and label generation unit, the performance indicators of candidate strategy parameters are statistically analyzed, and optimization algorithms are used to obtain strategy parameter sets of various styles.

[0029] 4) Model Library Management Unit: Maintain the mapping relationship between each type of working condition label combination and multiple styles of strategy parameter sets, and record the version number and update time corresponding to each mapping relationship to build a strategy parameter model library.

[0030] 5) Strategy parameter matching and distribution unit: In response to the request from the vehicle-side kinetic energy recovery control device of the target new energy micro-truck, based on the current operating condition label vector L and preference information, the device performs operating condition indexing and matching from the strategy parameter model library, selects or generates several sets of candidate strategies as subsets of the strategy parameter model library, packages them, and sends them to the vehicle-side kinetic energy recovery control device of the target new energy micro-truck.

[0031] (iii) User interaction and strategy selection device, which runs on the vehicle display screen or mobile terminal, is used to display the candidate strategy summary (estimated energy consumption, smoothness, safety margin, etc.) in the strategy parameter model library subset, receive user / fleet strategy selection, select a set of candidate strategies from the strategy parameter model library subset based on the user / fleet strategy selection as the strategy parameter set S_active, and deliver the strategy parameter set S_active to the vehicle-side kinetic energy recovery control device.

[0032] The second aspect of this invention discloses a data- and operating condition-based method for controlling the kinetic energy recovery of new energy micro-trucks, such as... Figure 1 As shown, it includes the vehicle-side portion of the vehicle-side kinetic energy recovery control device and the cloud portion of the cloud-side strategy generation device.

[0033] 1) Vehicle-side components (operational data acquisition, reporting, and real-time control) Step 101, Historical Data Collection and Reporting, further includes: Step 1011: Collect cruise data of the new energy micro-truck Throughout the entire operation of the new energy micro-truck, cruise data is collected by the data acquisition unit at a sampling period Δt (e.g., 100 ms–1 s) and cached locally.

[0034] When collecting cruise data, if a braking event (e.g., brake pedal travel s_brk exceeds the travel threshold s_th) or a significant deceleration event (e.g., longitudinal deceleration a < speed threshold a_th) is detected, the start and end times of the corresponding set of data are additionally recorded and marked as "critical deceleration segments". The daily operation of new energy micro-trucks generates a large amount of ordinary cruise data, but this data is not very valuable for calibrating kinetic energy recovery strategies. What truly affects energy recovery and safety are the periods of significant deceleration / braking. This invention marks these periods as "critical deceleration segments," allowing for focused analysis only of data from these segments during subsequent local or cloud-based statistical, modeling, and strategy optimization processes, thus reducing the amount of invalid data.

[0035] Step 1012, Trigger Upload If any of the following conditions are met, data reporting will be triggered, and the vehicle-side kinetic energy recovery control device will upload the data to the cloud-based strategy generation device: Condition 1) Upon completion of a single delivery task, it can be identified based on the route identifier (route_id) and mileage increment; Condition 2) Cumulative operating mileage ΔD ≥ mileage threshold D_th, for example, 50 km; Condition 3) The change in SOC |ΔSOC| ≥ the SOC threshold SOC_th, for example, 10%; Condition 4) The amount of locally cached data exceeds the capacity threshold.

[0036] Step 1013: Data Packaging and Uploading The cruise data collected and stored locally in step 1011 is packaged into a data packet P in chronological order. Each piece of data in data packet P includes a timestamp (time), route identifier (route_id), driver identifier (driver_id), vehicle speed (v), longitudinal deceleration (a), estimated vehicle mass or load (M_est), road gradient (θ), estimated road surface adhesion coefficient (φ_est), battery state of charge (SOC), battery temperature (T_bat), battery acceptable power (P_chg_max), current motor recovery torque (T_rec_cur), brake pedal travel (s_brk), braking duration (t_brk), and braking frequency (f_brk), etc.

[0037] The data packet P is compressed and encrypted, and the vehicle ID, software version number, and data batch number are added. It is then uploaded to the cloud-based policy generation device through the communication unit.

[0038] Step 102, Real-time Vehicle Operating Condition Identification and Policy Request, further includes: Step 1021: Real-time operating condition identification Within each control cycle, the operating condition identification unit calculates operating condition labels for each dimension based on the real-time cruise data collected by the data acquisition unit according to preset rules, and forms the current operating condition label vector L_cur={L_load,L_grade,L_mu,L_state, L_style,L_SOC, route_id}.

[0039] Step 1022: Request candidate strategies from the cloud. The current operating condition label vector L_cur, vehicle identifier vehicle_id, route identifier route_id, driver identifier driver_id, and user preset preferences (such as "default energy saving priority") are packaged into a request message Req and sent to the cloud policy generation device through the communication unit. This message is used to perform operating condition indexing and policy parameter requests in the pre-built policy parameter model library.

[0040] Step 103, Vehicle-side loading strategy and execution control, further includes the following steps; Step 1031: Receive candidate strategy parameter set Receive N sets of candidate policy parameter sets S_k, where N≥2, k=1…N, sent by the cloud-based policy generation device. Each set of candidate policy parameter sets S_k contains: Vehicle speed segment {v_i} and corresponding maximum regenerative torque limit T_rec_max(i); The rate of change of regenerative torque with respect to pedal travel / deceleration, k_rec(i); Low-speed exit threshold v_cut_low, SOC exit threshold SOC_cut; Slope compensation coefficient K_grade(θ), low adhesion reduction coefficient K_mu(φ_est); Mechanical braking compensation ratio α_hyd(i) and switching point s_hyd(i); And, strategic style identifiers (energy saving / smoothness / safety).

[0041] It also receives brief indicators calculated by the cloud-based strategy generation device, such as the expected energy saving rate, smoothness rating, and braking safety margin score.

[0042] Step 1032, User Selection The user interaction and strategy selection device displays the name and summary indicators of each candidate strategy in N sets of candidate strategy parameter sets S_k. The driver selects one of the candidate strategy parameter sets S_sel via touch screen / button. If no selection is made within a preset time, a candidate strategy parameter set S_sel is automatically selected based on brief indicators and priority (e.g., safety priority > smoothness > energy saving).

[0043] Step 1033, Control Execution The execution control unit loads the candidate policy parameter set S_sel as the policy parameter set S_active; For each deceleration event: The total required torque T_req is calculated based on the current vehicle speed v, the desired deceleration a_req, the brake pedal travel s_brk, and the strategy parameter set S_active. The desired deceleration a_req is determined using the following method: When there is an ACC / cruise control request, the expected deceleration a_req given by the ACC / cruise control module is used directly; When the driver presses the brake pedal, the desired deceleration a_req is obtained by looking up the pre-stored "paddle travel - desired deceleration" calibration table based on the current vehicle speed v and the brake pedal travel s_brk. The total required torque T_req is obtained using the following method: Based on the vehicle's longitudinal dynamics model, the desired deceleration a_req is converted into the total braking torque T_req required to achieve that deceleration, for example: T_req = r_w m_est a_req Where r_w is the effective radius of the wheel, and m_est is the estimated value of the vehicle mass or load.

[0044] Based on the segment to which the current vehicle speed v belongs, the gradient compensation coefficient K_grade(θ'), and the low adhesion degradation coefficient K_mu(φ_est'), the maximum recovery torque limit T_rec_max(i) is corrected to obtain the maximum recovery torque limit T_rec_max_corr under the current operating condition.

[0045] In this invention, the candidate strategy parameter set S_k obtained from the cloud only applies to "typical working conditions" and cannot cover all instantaneous road conditions. The cloud values ​​of the corresponding parameters in the candidate strategy parameter set S_k correspond to historical moments / typical values, including vehicle speed segments {v_i} and correction functions K_grade(θ) and K_mu(φ_est) that change with road slope θ and road surface adhesion coefficient estimation φ_est. Therefore, the maximum recovery torque limit T_rec_max_corr under the current working condition must be calculated online based on the real-time measured θ and φ_est in order to truly reflect the slope and adhesion status at the current moment and ensure safety and comfort.

[0046] In this invention, the correction of the maximum recovery torque upper limit T_rec_max(i) based on the segment to which the current vehicle speed v belongs and K_grade(θ) and K_mu(φ_est) includes the following steps: After obtaining the corresponding vehicle speed segment {v_i} based on the current vehicle speed v, read the upper limit of the baseline maximum recovery torque T_rec_max(i) corresponding to that segment.

[0047] Substitute the real-time measured road slope θ' into the slope compensation coefficient calculation function K_grade(θ) to obtain the current slope compensation coefficient K_grade(θ'); Substitute the real-time estimated road adhesion coefficient φ_est' into the low adhesion depreciation coefficient calculation function K_mu(φ_est) to obtain the current adhesion compensation coefficient K_mu(φ_est'); The maximum recoverable torque limit T_rec_max(i) is corrected by multiplication to obtain the maximum recoverable torque limit under the current operating condition, as shown in the following formula: T_rec_max_corr = T_rec_max(i) K_grade(θ') K_mu(φ_est') Calculate the expected recovery torque T_rec_cmd=min(T_req·K_rec,T_rec_max_corr,T_rec_bat(P_chg_max)); In the formula: T_rec_bat is calculated by the execution control unit based on the current battery state of charge (SOC), battery temperature (T_bat), battery receivable power (P_chg_max), and current vehicle speed (v) using a pre-calibrated "battery charging capacity – regenerative torque" mapping table, either by looking up the table or by interpolation. This mapping table is obtained through offline testing and simulation calibration, and comprehensively considers factors such as battery charging power limitations, motor / transmission efficiency, and constraints at different vehicle speeds. Therefore, T_rec_bat represents the maximum kinetic energy recovery torque that the vehicle can apply without exceeding the current battery charging capacity and safety constraints.

[0048] Remaining torque T_hyd_cmd=T_req T_rec_cmd is handled by the hydraulic brake; Motor regeneration and hydraulic braking are performed via the VCU and ESP modules, respectively.

[0049] Step 104, Security Constraints and Degradation Control If detected: The road surface adhesion coefficient φ_est is lower than the adhesion threshold φ_low (low adhesion road surface); Alternatively, the battery temperature T_bat exceeds the safe range [T_min, T_max]; Alternatively, the battery's receiveable power P_chg_max decreases significantly, meaning the current battery's receiveable power P_chg_max is less than the preset power lower limit P_min, where P_min is a parameter of the vehicle's motor. Alternatively, a fault indicator may appear in the braking system / critical sensors; Or, the communication unit is unable to connect to the cloud for an extended period of time; The security management unit implements the following policies: Lower the maximum regenerative torque limit T_rec_max_corr under the current operating conditions (e.g., scale it proportionally by K_safe<1), and increase the low-speed exit threshold v_cut_low to allow regeneration to exit earlier; If necessary, ignore the policy parameter set S_active and directly switch to the local basic safety policy S_safe_base (a pre-defined conservative policy with low recovery torque and high mechanical braking ratio). In this invention, the condition for switching the basic security policy S_safe_base is as follows: The road surface adhesion coefficient φ_est is lower than the adhesion threshold φ_low; Alternatively, the battery temperature T_bat exceeds the safe range [T_min, T_max]; Alternatively, the current battery's receiveable power P_chg_max is less than the preset power lower limit P_min.

[0050] When conditions return to normal, either restore to the policy parameter set S_active by annealing or request a new policy.

[0051] (ii) Cloud-based component (data processing, operational condition clustering, and strategy modeling) Step 201: Data Preprocessing and Sample Construction Perform the following operations on each received packet of historical data: Time alignment and missing value imputation; Outlier removal (such as outliers caused by sensor failure or communication packet loss); The data is divided into multiple running task samples based on route identifier (route_id), driver identifier (driver_id), date, and task segment.

[0052] Step 202: Working Condition Clustering and Label Generation Select the feature vector X = {average load ratio λ_load, slope statistics θ_mean / θ_max, speed distribution parameters, start-stop frequency f_start, road surface adhesion statistics φ_est, SOC distribution, braking frequency, etc.}; On the same route / driver, clustering algorithms (such as K-Means, hierarchical clustering) or rule-based binning methods are used to divide the samples into K typical working conditions; Assign a working condition label L={L_load,L_grade,L_mu,L_state,L_style,L_SOC,route_id} to each typical working condition, and record the corresponding threshold definition for online identification by the vehicle.

[0053] Step 203: Strategy Parameter Modeling and Optimization For each working condition label combination L_j, perform the following operations: Select existing historical strategy configurations (including original calibration strategies and experimental strategies) under the current operating conditions. Calculate the following performance metrics: Average recovered energy E_rec(j); Energy consumption per unit distance E_perkm(j); Tire slippage / ABS intervention frequency f_abs(j); Ride deceleration fluctuation σ_a(j) (ride comfort index); Emergency braking stopping distance / success rate P_safe(j); The policy parameter vector is denoted as: S={T_rec_max(i),k_rec(i),v_cut_low,SOC_cut,K_grade,K_mu,α_hyd(i),s_hyd(i)}; Under the constraints: f_abs(j) ≤ f_abs_max; And, P_safe(j) ≥ P_safe_min; And, σ_a(j) ≤ σ_a_max; And, the battery charging power constraint is not violated; In such cases, multi-objective optimization or heuristic search (such as grid search / genetic algorithm) can be used to find several Pareto optimal solutions; These solutions are clustered or sorted according to preferences and labeled as "Energy Saving Priority S_e(j)", "Smoothness Priority S_s(j)", "Safety Priority S_a(j)", etc. Store {L_j : S_e(j), S_s(j), S_a(j)…} in the model library.

[0054] Step 204: Candidate Strategy Generation and Distribution The cloud-based policy generation device received a request (Req) from the vehicle: Match the label L_j that is closest to L_cur (either a complete match or a nearest neighbor match can be used). Read the several strategies S_k(j) corresponding to L_j; Fine-tune parameters such as K_grade and K_mu based on the current environment (e.g., weather / road hazard information); Calculate the summary metrics (expected energy saving rate, smoothness level, safety margin) for each strategy and package and distribute them.

[0055] Those skilled in the art will understand that the features described in the various embodiments and / or claims of this invention can be combined or combined in various ways, even if such combinations or combinations are not explicitly described in this disclosure. In particular, the features described in the various embodiments and / or claims of this invention can be combined or combined in various ways without departing from the spirit and teachings of this disclosure. All such combinations and / or combinations fall within the scope of this disclosure.

[0056] The embodiments of the present invention have been described above. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of this disclosure. Although various embodiments have been described above, this does not mean that the measures in the various embodiments cannot be used advantageously in combination. The scope of the present invention is defined by the appended claims and their equivalents. Various substitutions and modifications can be made by those skilled in the art without departing from the scope of the present invention, and all such substitutions and modifications should fall within the scope of this disclosure.

Claims

1. A new energy micro-truck kinetic energy recovery control system based on data and operating conditions, characterized in that, Adopt vehicle end cloud The user interface has a three-layer architecture, including a vehicle-side kinetic energy recovery control device, a cloud-based strategy generation device, and a user interaction and strategy selection device, wherein: The vehicle-side kinetic energy recovery control device further includes: The data acquisition unit is used to collect cruise data of the target new energy micro-truck; The condition identification unit is used to calculate the load level label L_load, gradient level label L_grade, adhesion level label L_mu, driving state type label L_state, driving style level label L_style, and SOC interval label L_SOC based on the cruise data collected by the data acquisition unit, and generate the condition label vector L = {L_load, L_grade, L_mu, L_state, L_style, L_SOC, route_id}. The communication unit is used for bidirectional transmission of data and policy parameters with the cloud-based policy generation device. The policy storage unit is used to locally cache a subset of the policy parameter model library, the policy parameter set S_active, and the basic security policy S_safe_base issued from the cloud policy generation device; The execution control unit is used to calculate the expected total deceleration torque T_req, motor recovery torque T_rec_cmd, and hydraulic braking torque T_hyd_cmd based on the driver's braking requirements and strategy parameter set S_active of the target new energy micro-truck, and send the motor recovery torque T_rec_cmd to the vehicle controller or motor controller, and send the hydraulic braking torque T_hyd_cmd to the braking system. The safety management unit is used to monitor in real time the estimated value of road surface adhesion coefficient φ_est, battery temperature T_bat, battery receiveable power P_chg_max, braking system fault flags and communication status. Under the condition that the safety constraints are not met, it performs online derating or rollback to the basic safety policy S_safe_base for some parameters in the strategy parameter set S_active. The cloud-based policy generation device further includes: The data receiving and storage unit is used to receive historical operation data packets from multiple new energy micro-trucks. The historical operation data packets contain all the cruise data of each new energy micro-truck collected by the data acquisition unit. The working condition clustering and label generation unit is used to extract features and cluster historical sample data received by the data receiving and storage unit, generate a discrete set of working condition labels, and maintain label definitions and thresholds. The strategy parameter modeling unit is used to statistically analyze the performance indicators of candidate strategy parameters under each combination of working condition labels generated by the working condition clustering and label generation unit, and to obtain strategy parameter sets of various styles using optimization algorithms. The model library management unit is used to maintain the mapping relationship between each type of working condition label combination and multiple styles of strategy parameter sets, and to build and obtain the strategy parameter model library. The strategy parameter matching and distribution unit is used to respond to the request of the vehicle-side kinetic energy recovery control device of the target new energy micro-truck. Based on the current working condition label vector L_cur and preference information, it performs working condition indexing and matching from the strategy parameter model library, selects or generates several sets of candidate strategies as subsets of the strategy parameter model library, packages them and distributes them to the vehicle-side kinetic energy recovery control device of the target new energy micro-truck. The user interaction and strategy selection device is used to display the candidate strategy summary in the strategy parameter model library subset, receive user / fleet strategy selection, select a set of candidate strategies from the strategy parameter model library subset based on the user / fleet strategy selection as the strategy parameter set S_active, and deliver the strategy parameter set S_active to the vehicle-side kinetic energy recovery control device.

2. The new energy micro-truck kinetic energy recovery control system based on data and operating conditions as described in claim 1, characterized in that, The cruise data includes vehicle speed v, longitudinal deceleration a, wheel speed ω, gear / drive mode m, road gradient θ, ambient temperature T_env, road surface adhesion coefficient estimate φ_est, vehicle mass or load estimate M_est, battery state of charge (SOC), battery temperature T_bat, battery receivable power P_chg_max, current motor regenerative torque T_rec_cur, brake pedal travel s_brk, braking duration t_brk, braking frequency f_brk, whether emergency braking is required (flag_emg), route identifier (route_id), mileage location d, driver identifier (driver_id), and timestamp (time).

3. A new energy micro-truck kinetic energy recovery control system based on data and operating conditions as described in claim 2, characterized in that, Based on the estimated vehicle mass or load value M_est and the preset curb weight M_0, calculate the load ratio = (M_est - M_0) / M_0, and according to the interval of the calculated load ratio, discretize the current load level into at least one of light load, medium load and heavy load to obtain the load level label L_load. Based on the average value of the collected road slope θ within a preset time window, the current slope is divided into at least one of the following: flat road, uphill or downhill, and optional long downhill, to obtain the slope grade label L_grade. Based on the collected road surface adhesion coefficient estimate φ_est, the current adhesion level is divided into at least one of high adhesion, medium adhesion and low adhesion according to the preset adhesion coefficient range, so as to obtain the adhesion level label L_mu. Within a preset time or mileage window, the vehicle speed v, longitudinal deceleration a, braking duration t_brk, and braking frequency f_brk are statistically analyzed. Based on the average vehicle speed, the degree of vehicle speed fluctuation, and the braking frequency per unit mileage, the current driving state is divided into at least one of frequent start-stop, long downhill, cruise, or other mixed conditions to obtain the driving state type label L_state. Based on the longitudinal acceleration and deceleration characteristics, accelerator pedal usage characteristics, and braking operation characteristics of the driver identification within a preset mileage range, a driving style score is calculated, and the current driving style is divided into at least one of economic / mild, normal, or aggressive according to the interval in which the driving style score is located, so as to obtain the driving style level label L_style. Based on the battery state of charge (SOC), battery temperature (T_bat), and battery power (P_chg_max), the current SOC state is divided into at least one of a low SOC range, a medium SOC range, or a high SOC range to obtain the SOC range label L_SOC.

4. A data- and operating condition-based kinetic energy recovery control method for new energy micro-trucks, employing the new energy micro-truck kinetic energy recovery control system described in claim 1, characterized in that... This includes the vehicle-side portion of the vehicle-side kinetic energy recovery control device and the cloud-side portion of the cloud-based strategy generation device, wherein: The vehicle-end portion includes the following steps: Step 1: Collect cruise data of new energy micro-trucks, package the collected cruise data and upload it to the cloud strategy generation device; Step 2: Within each control cycle, the operating condition identification unit forms the current operating condition label vector L_cur={L_load,L_grade,L_mu,L_state, L_style,L_SOC,route_id} based on the real-time cruise data collected by the data acquisition unit. Then, it packages the previous operating condition label vector L_cur together with the vehicle identifier vehicle_id, route identifier route_id, driver identifier driver_id, and user preset preferences into a request message Req and sends it to the cloud policy generation device. Step 3: The vehicle-side kinetic energy recovery control device receives N sets of candidate strategy parameters S_k, where N≥2, k=1…N, fed back by the cloud-based strategy generation device; Step 4: The user selects one set of candidate policy parameters from N sets of candidate policy parameter sets S_k through the user interaction and policy selection device as the policy parameter set S_active; Step 5: The execution control unit loads the strategy parameter set S_active, calculates the expected total deceleration torque T_req, motor recovery torque T_rec_cmd, and hydraulic braking torque T_hyd_cmd, and sends the motor recovery torque T_rec_cmd to the vehicle controller or motor controller, and sends the hydraulic braking torque T_hyd_cmd to the braking system. Step 6: Monitor the estimated road surface adhesion coefficient φ_est, battery temperature T_bat, battery receiveable power P_chg_max, braking system fault flags, and communication status in real time. If safety constraints are not met, derate some parameters in the strategy parameter set S_active online or roll back to the basic safety strategy S_safe_base. After returning to normal, restore to the strategy parameter set S_active by annealing or request a new strategy. The cloud component includes the following steps: Step A: The cloud-based strategy generation device collects historical operation data packets from multiple new energy micro-trucks to construct operation task samples; Step B: Based on the extracted feature vector X={average load ratio λ_load, gradient statistics θ_mean / θ_max, speed distribution parameters, start-stop frequency f_start, road surface adhesion statistics φ_est, SOC distribution, braking frequency}, for the same route or driver, the running task samples are divided into K typical working conditions, and the working condition label vector L={L_load,L_grade,L_mu,L_state,L_style,L_SOC,route_id} is obtained for each typical working condition. Step C: For each combination of working conditions, use multi-objective optimization or heuristic search to find several Pareto optimal solutions for all historical strategy configurations according to budget constraints, and cluster these Pareto optimal solutions according to preferences to form different styles of strategy parameter sets, and then store them in the strategy parameter model library. Step D: After receiving the request message Req, the cloud-based strategy generation device matches the current operating condition label vector L_cur with the operating condition label vector in the strategy parameter model library, and matches the preference information with different styles in the strategy parameter model library. The obtained matched strategy parameter set is then sent to the vehicle-side kinetic energy recovery control device as a subset of the strategy parameter model library.

5. The energy recovery control method for new energy micro-trucks based on data and operating conditions as described in claim 4, characterized in that, In step 1, when collecting cruise data of new energy micro-trucks, if a braking event or significant deceleration event is detected, the start and end times of the corresponding cruise data are recorded, and the current set of cruise data is marked as "critical deceleration segment". Only the cruise data marked as "critical deceleration segment" is packaged and uploaded. A significant deceleration event refers to an event in which the longitudinal deceleration a < speed threshold a_th.

6. The energy recovery control method for new energy micro-trucks based on data and operating conditions as described in claim 4, characterized in that, In step 1, the packaging and uploading of cruise data is triggered only if any of the following conditions are met: Condition 1) The current single delivery task of the new energy micro-truck has ended; Condition 2) Cumulative operating mileage ΔD ≥ mileage threshold D_th; Condition 3) The change in SOC |ΔSOC| ≥ the SOC threshold SOC_th; Condition 4) The amount of locally cached data exceeds the capacity threshold.

7. The energy recovery control method for new energy micro-trucks based on data and operating conditions as described in claim 4, characterized in that, In step 3, each set of candidate policy parameter sets S_k includes: Vehicle speed segment {v_i} and corresponding maximum regenerative torque limit T_rec_max(i); The rate of change of regenerative torque with respect to pedal travel / deceleration, k_rec(i); Low-speed exit threshold v_cut_low, SOC exit threshold SOC_cut; The slope compensation coefficient calculation function is K_grade(θ), and the low adhesion reduction coefficient calculation function is K_mu(φ_est). Mechanical braking compensation ratio α_hyd(i) and switching point s_hyd(i); And, strategic style identifiers.

8. The energy recovery control method for new energy micro-trucks based on data and operating conditions as described in claim 4, characterized in that, In step 3, the vehicle-side kinetic energy recovery control device also receives brief indicators; In step 4, if the user does not make a selection, a candidate strategy parameter set is automatically selected from the N candidate strategy parameter sets S_k according to the pre-set priority based on the brief indicators.

9. A data- and operating condition-based kinetic energy recovery control method for new energy micro-trucks as described in claim 7, characterized in that, In step 5, the desired total deceleration torque T_req is calculated using the following formula: T_req = r_w m_est a_req Where r_w is the effective radius of the wheel, m_est is the estimated mass or load of the vehicle, and a_req is the expected deceleration; The motor recovery torque T_rec_cmd is calculated using the following formula: T_rec_cmd=min(T_req·k_rec(i),T_rec_max_corr,T_rec_bat(P_chg_max)); In the formula: T_rec_max_corr is the upper limit of the maximum recovery torque under the current operating conditions, T_rec_max_corr = T_rec_max(i) K_grade(θ') K_mu(φ_est') substitutes the real-time measured road slope θ' into the slope compensation coefficient calculation function K_grade(θ) to obtain the current slope compensation coefficient K_grade(θ'). K_mu(φ_est') substitutes the real-time estimated road surface adhesion coefficient φ_est' into the low adhesion depreciation coefficient calculation function K_mu(φ_est) to obtain the current adhesion compensation coefficient K_mu(φ_est'). T_rec_bat(P_chg_max) is the maximum regenerative torque allowed under the current operating conditions. It is obtained by the execution control unit through table lookup or interpolation calculation in a pre-calibrated "battery charging capacity - regenerative torque" mapping table based on the current battery state of charge (SOC), battery temperature (T_bat), battery receptive power (P_chg_max), and current vehicle speed (v). The "battery charging capacity - regenerative torque" mapping table is obtained through offline testing and simulation calibration. The hydraulic braking torque T_hyd_cmd is calculated using the following formula: T_hyd_cmd=T_req T_rec_cmd.

10. The energy recovery control method for new energy micro-trucks based on data and operating conditions as described in claim 4, characterized in that, In step 6, the conditions under which the security constraints are not met include: The road surface adhesion coefficient φ_est is lower than the adhesion threshold φ_low; Or, the battery temperature T_bat exceeds the preset safety range; Alternatively, the current battery's receiveable power P_chg_max is less than the preset power lower limit P_min; Alternatively, a fault indicator may appear in the braking system / critical sensors; Alternatively, the communication unit may be unable to connect to the cloud for an extended period of time.

Images