A Safety Management Method for the Three-Electric Systems of New Energy Engineering Vehicles

Abstract

This invention relates to the field of new energy vehicle electric drive technology and discloses a safety management method for the electric drive system of new energy vehicles. It involves synchronously collecting multi-dimensional data from the battery, motor, and electronic control systems of the vehicle, performing real-time feature extraction and analysis on the vehicle to obtain early warning coefficients for each subsystem. If any coefficient exceeds a safety threshold, a local edge warning is immediately triggered. For vehicles that do not trigger a warning, a warning index reflecting their overall health status is calculated through trend analysis based on their historical status data. Simultaneously, based on the electronic layout map of the construction site and risk attributes, a site-wide operational risk overlay map is generated through spatial overlay calculation. Finally, using the vehicle warning index and operational risk value as input, and aiming to maximize safety adaptability, global optimization matching is performed under various constraints to generate a vehicle task allocation scheme. This scheme is dynamically adjusted based on task completion status or sudden warnings to achieve collaborative safety management.

Description

Technical Field

[0001] This invention relates to the field of new energy vehicle electric drive technology, specifically a safety management method for new energy vehicle electric drive systems. Background Technology

[0002] With the increasing urgency of global energy structure adjustment and environmental protection, the application of new energy technologies in the field of construction machinery has become an irreversible trend. Construction vehicles (such as excavators, loaders, cranes, mining trucks, etc.) are gradually transforming from traditional internal combustion engine drive to electric drive. Among them, the battery, motor, and electronic control (collectively known as the three-electric system) are the core of new energy vehicles, and their performance and safety directly determine the reliability, working efficiency, and commercial value of the entire vehicle. Therefore, the new energy three-electric system is very important for construction vehicles. Engineering vehicles typically operate in high-intensity, high-load, and continuous vibration environments, often facing harsh conditions such as high dust, high humidity, and significant temperature fluctuations. This poses a significant challenge to the mechanical strength and thermal management capabilities of battery packs. Furthermore, the high-voltage circuits (typically reaching hundreds of volts) in the three-electric system pose a risk of electric shock. Overcharging, over-discharging, short circuits, or mechanical abuse of the power battery can easily lead to thermal runaway, resulting in fires or even explosions, with particularly severe consequences in enclosed or high-risk work sites (such as mines and tunnels). Faced with these dangers, traditional safety monitoring and management often remain at the alarm stage after a fault occurs, lacking predictive maintenance guidance based on the actual health status of the equipment. Moreover, vehicle operating data and fault data are often isolated from each other. However, such decentralized protection strategies often fail to achieve coordinated safety management of the overall power system when dealing with the complex coupled operating conditions of engineering vehicles. Summary of the Invention

[0003] The purpose of this invention is to solve the problems mentioned in the background art by proposing a safety management method for the three-electric system of new energy engineering vehicles.

[0004] The objective of this invention can be achieved through the following technical solution: a safety management method for the three-electric system of new energy engineering vehicles, comprising the following steps: Step 1: Synchronously collect data on the three key electrical components (battery, motor, and electronic control). Battery data includes the individual cell voltage and temperature, as well as the battery pack's State of Charge (SOC) and State of Health (SOH) (characterizing the degree of degradation of maximum usable capacity, usually expressed as a percentage). Motor data includes rotor speed, actual output torque, motor temperature, and coolant temperature. Electronic control data includes controller temperature, packet loss rate, and latency in data transmission. Step 2: Based on the data of the three electrical components, perform edge feature extraction and analysis to obtain the warning coefficients of the battery, motor, and electronic control. If the warning coefficient of the battery, motor, or electronic control exceeds its corresponding threshold, an edge warning is triggered; otherwise, proceed to step 3. Step 3: For vehicles that have not triggered edge warnings, perform comprehensive state analysis based on all warning coefficients of their historical status to obtain the warning index; Step 4: Based on the electronic layout map of the construction site and the attributes of the work site, conduct site risk overlay analysis and output the work risk overlay map; Step 5: Based on the warning index and the operational risk overlay map, all vehicles that have not triggered edge warnings are globally optimized and allocated.

[0005] In a preferred embodiment of the present invention, step two specifically includes: S2-1, Edge feature extraction and analysis are performed based on battery data to obtain the battery early warning coefficient; S2-2, Edge feature extraction and analysis are performed based on motor data to obtain the motor early warning coefficient; S2-3, Edge feature extraction and analysis are performed based on electronic control data to obtain the electronic control early warning coefficient; S2-4, with preset battery warning threshold, motor warning threshold, and electronic control warning threshold, respectively, compares the battery warning coefficient, motor warning coefficient, and electronic control warning coefficient with the battery warning threshold, motor warning threshold, and electronic control warning threshold, respectively. If the battery warning coefficient > the battery warning threshold, a battery warning is triggered; if the motor warning coefficient > the motor warning threshold, a motor warning is triggered; if the electronic control warning coefficient > the electronic control warning threshold, an electronic control warning is triggered. The edge warning includes battery warning, motor warning, and electronic control warning.

[0006] In a preferred embodiment of the present invention, step S2-1 specifically includes: The individual cell voltages of all cells in the battery pack are calculated using the standard deviation formula to determine the individual cell voltage consistency, thus obtaining the individual cell voltage coefficient. The temperature uniformity of all individual cells in the battery pack is calculated using the standard deviation formula, and the individual cell temperature coefficient is obtained. Preset with a safe minimum remaining battery power (SOC) safe If the remaining state of charge (SOC) of the battery is greater than or equal to the safe minimum SOC of the remaining charge... safe Then the remaining power coefficient is set to zero, i.e., R SOC =0; otherwise, the remaining power coefficient is: ;where clip(*) is the truncation function; The battery state of health (SOH) is determined using a cutoff formula. The degradation state coefficient R was calculated.SOH ; The number of battery error reports between the last monitoring time and the current monitoring time is recorded as m. 电池 and through the truncation formula Normalization is performed to obtain the battery error coefficient R. 电池 ; This represents the maximum number of errors allowed within the monitoring period. A weighting factor is assigned to each of the cell voltage coefficient, cell temperature coefficient, remaining capacity coefficient, degradation state coefficient, and battery error coefficient. The cell voltage coefficient, cell temperature coefficient, remaining capacity coefficient, degradation state coefficient, and battery error coefficient are then fused by linear weighting to obtain the battery warning coefficient.

[0007] In a preferred embodiment of the present invention, step S2-2 specifically includes: The motor rotor speed n is obtained through the formula Convert the angular velocity to ω, and then multiply the angular velocity by the actual output torque of the motor to obtain the mechanical power P. out Through engineering loss model The copper loss, iron loss, and mechanical loss were calculated separately; where kt is the motor torque constant. This indicates the motor temperature T motor The phase resistance is given by: a is the hysteresis loss coefficient, b is the eddy current loss coefficient, and c is the mechanical loss coefficient. The heat generation rate Q is then calculated by adding the copper loss, iron loss, and mechanical loss. gen Then, based on the thermal model The overheating risk coefficient λ is obtained; where, For the motor thermal resistance, For the heat capacity of the motor, T cool This refers to the coolant temperature. The number of times the motor reported an error between the last monitoring time and the current monitoring time is recorded as m. 电机 and through the truncation formula Normalization is performed to obtain the motor error coefficient R. 电机 ; A weighting factor is assigned to both the overheating risk coefficient and the motor error coefficient, and these two factors are then fused together using linear weighting to obtain the motor early warning coefficient.

[0008] In a preferred embodiment of the present invention, steps S2-3 specifically include: The temperature T of the electronically controlled controller contorl Using S-shaped response function Simulate the linear relationship β and calculate the control risk coefficient β, T criticalis the critical temperature threshold of the controller chip in the electronic control system, which represents the highest operating temperature of the controller; h is the curve steepness factor, used to control the degree of drastic transition from low to high risk. The packet loss rate and latency rate during data transmission in the electronic control system are normalized. Specifically, the normalization process involves dividing the packet loss rate and latency rate by the historical highest packet loss rate and latency rate of the electronic control system, respectively. Then, the normalized packet loss rate D and latency rate Y are processed using a parallel failure model. The communication risk coefficient γ is calculated. The number of times the electronic control system reported errors between the last monitoring time and the current monitoring time is recorded as m. 电控 and through the truncation formula Normalization is performed to obtain the electronic control error coefficient R. 电控 ; A weighting factor is assigned to each of the control risk coefficient, communication risk coefficient, and electrical control error coefficient. The control risk coefficient, communication risk coefficient, and electrical control error coefficient are then fused together by linear weighting to obtain the electrical control early warning coefficient.

[0009] In a preferred embodiment of the present invention, step three specifically includes: The system acquires the battery warning coefficient, motor warning coefficient, and electronic control warning coefficient for each historical monitoring of the vehicle, assigns a weight factor to each of these coefficients, and then performs a linear weighted fusion of the three coefficients to obtain the warning status value. This allows the system to track the warning status value corresponding to each vehicle monitoring session. A two-dimensional rectangular coordinate system is constructed with monitoring time as the horizontal axis and warning status value as the vertical axis. The warning status values ​​of each monitoring session are input into the coordinate system according to the corresponding monitoring time to form several discrete points, and then connected by a smooth curve to obtain the warning status value curve. By smoothing the warning status value curve and mining the tangent slope trend, a warning index q is generated for vehicles that have not triggered edge warnings. From this, the warning index q for all vehicles that have not triggered alarms can be obtained. r And pass it to step five; where r is the index of the vehicle that did not trigger the edge warning.

[0010] In a preferred embodiment of the present invention, the method for generating the warning index for vehicles that have not triggered edge warning specifically includes: Tangent lines are drawn to the curve at each discrete point in the warning status value curve, and the slope of the tangent lines is calculated using the least squares method. The increasing trend of the warning status value is calculated by averaging all slopes greater than zero, denoted as f1. Simultaneously, the decreasing trend of the warning status value is calculated by averaging the absolute values ​​of slopes less than zero, denoted as f2. Next, the warning status values ​​at each historical monitoring time are averaged to obtain the warning baseline value, denoted as F. The increasing trend f1, decreasing trend f2, and warning baseline value F are then amplified using a trend amplification formula. The early warning index q is calculated.

[0011] In a preferred embodiment of the present invention, the method for generating the operation risk overlay map in step four includes: An electronic layout map of the entire construction site is pre-stored. This map marks various work areas requiring engineering vehicle operations and their associated work attributes. Each work attribute is pre-defined to correspond to a hazard type value and a reduction coefficient. The work attribute of each work area is compared with all pre-defined work attributes to match its corresponding hazard type value, denoted as H. j Where j is the index of any work site; several points are evenly set on the electronic layout map, where any one point is denoted as g, and g is other points on the electronic layout map that do not include the work site; The distance d between each work site and point g is calculated using the Euclidean spatial distance formula. j (g), the specific formula is as follows: Wherein, the three-dimensional coordinates of point g are (x(g), y(g), z(g)), and the three-dimensional coordinates of work site j are (x(g), y(g), z(g)) j y j , z j ); Take any point g and assign the hazard type value H to the work site. j and attenuation coefficient K j Through formula The risks at each point in the electronic layout map are quantified and superimposed to obtain the risk superposition value for each point; the risk superposition value of each point is marked on the electronic partial map to obtain the operation risk superposition map.

[0012] In a preferred embodiment of the present invention, step five specifically includes: S5-1: Obtain the work route for each work site, identify the points on the work route and their corresponding risk superposition values, and calculate the route risk value by averaging the risk superposition values ​​of all points on the work route, denoted as L. j Risk type value H for the work site j Its corresponding line risk value L jAssign a weighting factor and assign the risk type value H j and line risk value L j A linear weighted fusion calculation is performed to sum the risks encountered when going to the work site, resulting in a work risk value. The work risk value for each work site is denoted as G. j ; S5-2, Establish a negative correlation between vehicle status and operational risk, i.e., the fit calculation formula: J rj This represents the fit between vehicle r and work site j. Qmin and Qmax refer to the minimum and maximum values ​​of the warning index among all vehicles that have not triggered the edge warning. Gmin and Gmax refer to the minimum and maximum values ​​of the work risk value among all work sites. Constraint settings: (1) Hard safety constraints: When the vehicle's warning index q r Exceeding the warning safety threshold q safe At that time, it is forbidden to be assigned to high-risk work sites, when G j >G safe (2) Operational capacity constraints: Ensure that the vehicle type matches the task requirements of the operation site; Based on the fit degree calculation and constraint conditions, output the fit degree matrix J for all vehicle-work site combinations; S5-3, Global Optimization Matching Algorithm: Construct an optimization objective function with the goal of maximizing overall fit: , where X rj X is the decision variable, including 0 and 1. rj =1 indicates that vehicle r is assigned to work site j, X rj =0 indicates that vehicle r is not assigned to work location j; the following conditions must be met during the global optimization process: (1) At least one vehicle shall be assigned to each work site: ΣX rj ≥1, j; (2) Each vehicle can be assigned to a maximum of one work site: ΣX rj ≤1, r; (3) ΣX rj ≤W j , j, W j The maximum vehicle capacity of work site j; According to fit J rj Sort all possible vehicle-work site pairs from high to low and assign them sequentially. If the constraints are met, confirm the assignment until all work sites have vehicles assigned or no vehicles can be assigned. Then output the optimal assignment scheme, that is, the work site assignment result for each engineering vehicle. S5-4, Safety Collaborative Management Execution: The allocation results are sent to each vehicle terminal via the vehicle network.

[0013] Compared with the prior art, the beneficial effects of the present invention are: 1. By extracting and analyzing features from the three electrical components (battery, motor, and electronic control system) data, early warning coefficients are obtained, and edge warnings are performed accordingly. This ensures the real-time nature of early warnings for engineering vehicles, effectively isolates primary risks, prevents them from escalating further, and enables rapid local identification and response to safety risks, forming the first line of defense against safety risks that does not rely on the network.

[0014] 2. By linearly fusing the warning coefficients of the three electric components, a time-series warning state value is obtained. A vehicle warning index is generated by using a smoothing method and trend mining of the tangent slope. This index can keenly capture the signs of vehicle performance deterioration and measure the current risk level of the vehicle.

[0015] 3. By introducing an exponential decay model based on physical laws, the propagation and superposition effects of risks of different natures in space are accurately quantified, generating a visualized operational risk superposition map. This can transform qualitative and local safety experience into a quantitative and global digital risk field, providing an intuitive and accurate spatial safety benchmark for global vehicle scheduling.

[0016] 4. By coupling vehicle warning status with work site risks at the global level, a negatively correlated fit model is established, and global optimization is performed under multiple safety and physical constraints to assign the healthiest vehicles to the most dangerous tasks. This significantly improves the operational safety and system robustness of the engineering vehicle convoy without reducing operational efficiency, and achieves safe collaborative management of engineering vehicles. Attached Figure Description

[0017] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0018] Figure 1 This is a general block diagram of the present invention. Detailed Implementation

[0019] Example Please see Figure 1 A safety management method for the three-electric system of new energy engineering vehicles, specifically including: Step 1: Comprehensive Data Sensing for All Three Electric Systems: This involves simultaneously collecting battery data, motor data, and electronic control data from the battery system, motor system, and electronic control system. Battery data includes the individual cell voltage and temperature, as well as the battery pack's State of Charge (SOC) and State of Health (SOH) (representing the degree of degradation of maximum usable capacity, usually expressed as a percentage). Motor data includes motor rotor speed, actual motor output torque, motor temperature, and coolant temperature. Electronic control data includes controller temperature, packet loss rate, and latency in data transmission. Step 2: Edge feature extraction and analysis for preliminary edge warning: S2-1, Battery edge feature extraction and analysis: Extract battery data, which includes cell voltage vi, cell temperature Ti, battery remaining state of charge (SOC), battery state of health (SOH), and battery error count. i represents the index of a single cell in the battery. The individual cell voltages vi of all cells in the battery pack are calculated using the standard deviation formula. Calculate the individual cell voltage consistency to obtain the individual cell voltage coefficient C1; i=1,2,3……I, where I is a positive integer and I represents the total number of individual cells in the battery pack; The individual cell temperature Ti of all cells in the battery pack is calculated using the standard deviation formula. The temperature uniformity of the monomer was calculated, and the monomer temperature coefficient C2 was obtained. Preset with a safe minimum remaining battery power (SOC) safe Remaining battery power safety lower limit SOC safe The setting is based on the electrochemical characteristics of lithium-ion batteries, aiming to prevent irreversible damage and safety risks caused by over-discharge, while also taking into account the high-power load requirements of engineering vehicles; if the remaining state of charge (SOC) of the battery is greater than or equal to the safe lower limit of remaining charge (SOC)... safe Then the remaining power coefficient is set to zero, i.e., R SOC =0; otherwise, the remaining power coefficient is: `clip(*)` is a truncation function that forcibly limits the remaining energy coefficient to the range of 0-1 to avoid negative or values ​​exceeding 1. In practical applications, the risk meaning of SOC varies depending on the scenario. Low SOC carries high risk during discharge and high SOC carries high risk during charging. However, since discharge is the primary mode of operation for conventional engineering vehicles, a low SOC risk model, i.e., the truncation function method, is used to calculate the remaining energy coefficient. The lower the SOC, the higher the risk of over-discharge, and thus the larger the remaining energy coefficient. A lower State of Health (SOH) indicates more severe battery degradation and a higher risk. Therefore, the truncation formula can be used directly. The degradation state coefficient R was calculated. SOH ; The number of battery error reports between the last monitoring time and the current monitoring time is recorded as m. 电池 and through the truncation formula Normalization is performed to obtain the battery error coefficient R. 电池 ; The maximum number of errors allowed within the monitoring period is set to 5 in practical applications, and can be modified according to actual needs. Battery errors include single cell voltage exceeding limits, single cell temperature exceeding limits, battery pack voltage exceeding limits, and battery pack temperature exceeding limits. For the individual cell voltage coefficient C1, individual cell temperature coefficient C2, and remaining charge coefficient R... SOC Degradation state coefficient R SOH and battery error coefficient R 电池 Each is assigned a weighting factor, and the individual cell voltage coefficient C1, individual cell temperature coefficient C2, and remaining energy coefficient R are linearly weighted. SOC Degradation state coefficient R SOH and battery error coefficient R 电池 The battery early warning coefficient is obtained by fusion; S2-2, Edge Feature Extraction and Analysis of Motor: Extracting motor data, including motor rotor speed n, actual motor output torque τ, and motor temperature T. motor Coolant temperature T cool The motor rotor speed is obtained through the formula Convert the angular velocity to ω, and then multiply the angular velocity by the actual output torque of the motor to obtain the mechanical power P. out Through engineering loss model The copper loss, iron loss, and mechanical loss were calculated separately; where k t The torque constant of the motor. This indicates the motor temperature T motor The phase resistance is given by: a is the hysteresis loss coefficient, b is the eddy current loss coefficient, and c is the mechanical loss coefficient. The heat generation rate Q is then calculated by adding the copper loss, iron loss, and mechanical loss. gen Then, based on the thermal model The overheating risk coefficient λ is obtained, where For the motor thermal resistance, For the heat capacity of the motor, T cool This refers to the coolant temperature. The number of times the motor reported an error between the last monitoring time and the current monitoring time is recorded as m. 电机 and through the truncation formula Normalization is performed to obtain the motor error coefficient R. 电机 ; To determine the maximum number of errors allowed within the monitoring period, in practical applications, motor errors include over-torque, over-speed, coolant temperature exceeding limits, and motor temperature exceeding limits. A weighting factor is assigned to the overheating risk coefficient and the motor error coefficient respectively, and then the overheating risk coefficient and the motor error coefficient are fused together by linear weighting to obtain the motor early warning coefficient; S2-3, Edge Feature Extraction and Analysis of Electronic Control: Extract the electronic control data, which includes the controller temperature T. contorl The packet loss rate (D) and latency (Y) in data transmission; the performance and lifespan of the controller chip in electronic control are strongly correlated with the controller temperature, and the risk of the controller is not linearly related to temperature, but increases sharply after a certain critical point, affecting the controller temperature (T). contorl Using S-shaped response function Simulate this linear relationship β and calculate the control risk coefficient β, T critical is the critical temperature threshold of the controller chip in the electronic control system, representing the highest operating temperature of the controller; h is the curve steepness factor, which represents the degree of drastic transition of control risk from low to high, and takes a value of 0.2; in this formula, when T contorl Much lower than T critical At that time, T contorl - critical When T is negative, the risk coefficient β approaches 0; contorl equal to T critical When T is zero, the control risk coefficient β equals 0.5; when T... contorl Much higher than T critica The exponent is a very large negative number, and the control risk coefficient β approaches 1; Packet loss rate and latency rate are two independent but jointly influential indicators of electronic control communication quality. A parallel failure model is used for their quantitative analysis, meaning that if either indicator deteriorates, the overall communication risk will significantly increase. First, the packet loss rate and latency rate are normalized by dividing each by the historical highest packet loss rate and latency rate in the electronic control system. Then, the normalized packet loss rate D and latency rate Y are analyzed using the parallel failure model. The communication risk coefficient γ is calculated. The number of times the electronic control system reported errors between the last monitoring time and the current monitoring time is recorded as m. 电控 and through the truncation formula Normalization is performed to obtain the electronic control error coefficient R. 电控 ; The maximum number of errors allowed within the monitoring period is specified. In practical applications, electronic control errors include communication errors, controller error restarts, and communication timeouts. A weighting factor is assigned to each of the control risk coefficient, communication risk coefficient, and electrical control error coefficient, and the electrical control early warning coefficient is obtained by merging the control risk coefficient β, communication risk coefficient, and electrical control error coefficient through linear weighting. S2-4 has preset warning thresholds for the battery, motor, and electronic control systems. These thresholds are set based on the inherent physical characteristics, material safety boundaries, and functional safety requirements of their respective subsystems. Specifically, the battery warning threshold is primarily determined based on the cell's electrochemical safety window, thermal runaway critical conditions, and historical performance degradation data. The motor warning threshold is mainly determined based on the heat resistance grade of the motor winding insulation material, the demagnetization temperature of the permanent magnet, and the junction temperature limits of the power devices. The electronic control warning threshold is primarily determined based on the semiconductor junction temperature limits of the control chip and the real-time reliability requirements of the vehicle bus communication protocol. Determine the required functional safety standards; compare the battery warning coefficient, motor warning coefficient, and electronic control warning coefficient with the battery warning threshold, motor warning threshold, and electronic control warning threshold, respectively. If the battery warning coefficient > the battery warning threshold, trigger the battery warning; if the motor warning coefficient > the motor warning threshold, trigger the motor warning; if the electronic control warning coefficient > the electronic control warning threshold, trigger the electronic control warning. Since the battery warning, motor warning, and electronic control warning are edge warnings for engineering vehicles, they are classified as edge warnings. If the vehicle does not trigger an edge warning, send the battery warning coefficient, motor warning coefficient, and electronic control warning coefficient to step three. By extracting and analyzing features from the three electrical components (battery, motor, and electronic control system) data, early warning coefficients are obtained, and edge warnings are performed accordingly. This ensures the real-time nature of early warnings for engineering vehicles, effectively isolates primary risks, prevents them from escalating further, and enables rapid local identification and response to safety risks, forming the first line of defense against safety risks that does not rely on the network.

[0020] Step 3: Perform comprehensive status analysis on vehicles that have not triggered edge warnings to obtain the warning index for each vehicle; specifically: The system acquires the battery warning coefficient, motor warning coefficient, and electronic control warning coefficient for each historical vehicle monitoring session, assigns a weighting factor to each coefficient, and then performs a linear weighted fusion to obtain the warning status value. This allows the system to track the warning status value for each vehicle monitoring session. A two-dimensional Cartesian coordinate system is constructed with monitoring time as the horizontal axis and warning status value as the vertical axis. The warning status values ​​for each historical monitoring session are input into the coordinate axis according to the corresponding monitoring time to form several discrete points, and then connected by a smooth curve to obtain a warning status value curve. Tangent lines are drawn to the curve at each discrete point in the warning status value curve, and the slope of the tangent lines is calculated using the least squares method. When the slope is greater than zero, it indicates that the warning status value shows an increasing trend at the discrete point; if the slope is less than zero, it indicates that the warning status value shows a decreasing trend at the discrete point. The increasing trend of the warning status value is calculated by averaging all slopes greater than zero and recorded as f1. At the same time, the decreasing trend of the warning status value is calculated by averaging the absolute values ​​of slopes less than zero and recorded as f2. Then, the warning status value at each historical monitoring time is averaged to obtain the warning baseline value, recorded as F. The increasing trend f1, decreasing trend f2, and warning baseline value F are then amplified using a trend amplification formula. The warning index q is calculated, and from this, the warning index q for all vehicles that have not triggered the edge warning can be obtained. r And pass it to step five; where r is the index of the vehicle that did not trigger the edge warning; By linearly fusing the warning coefficients of the three electric components, a time-series warning state value is obtained. A smoothing method and tangent slope trend mining are used to generate a warning index for vehicles that have not triggered edge warnings. This can keenly capture the signs of vehicle performance degradation and measure the current risk level of the vehicle.

[0021] Step 4: Perform site risk overlay analysis based on the vehicle operation site to output an operation risk overlay map: An electronic layout map of the entire construction site is pre-stored. The electronic partial map marks various work areas requiring engineering vehicle operations and their corresponding work attributes. In practical applications, attributes refer to the type of work area, such as enclosed attributes (tunnel interiors, pipeline corridors, underground utility tunnels, etc.), high-altitude attributes (elevated bridge edges, deep foundation pit edges, slopes, etc.), flammable and explosive attributes (temporary fuel tank areas, oxygen / acetylene cylinder storage points, paint and other chemical warehouses, asphalt heating areas, etc.), and high-voltage electrical attributes (surroundings of temporary substations / distribution boxes, areas under or adjacent to high-voltage transmission lines, areas under electrified railway contact networks, etc.). Those skilled in the art will understand that the work area attribute information listed above is only a partial example and not an exhaustive list. In practical applications, other work area attributes can be further defined and expanded according to specific construction scenarios and risk types. Each operational attribute is pre-defined with a corresponding hazard type value and attenuation coefficient. The hazard type value is set based on the inherent risk level and potential severity of accident consequences corresponding to different operational attributes. The assignment principle follows industry safety standards (such as GB 6441 "Classification of Enterprise Employee Injuries and Fatalities") and historical accident statistical analysis data, assigning higher values ​​to attributes that may lead to significant personnel casualties, equipment damage, or environmental destruction. The attenuation coefficient is set based on the spatial propagation and attenuation laws of the risk physical field. The principle is that the impact of risk on the surrounding environment weakens with increasing distance, but different types of risks have different attenuation rates. The determination of the coefficient depends on the simulation calculation of specific risk models (such as fire thermal radiation intensity attenuation models and explosion overpressure attenuation models) and the fitting of on-site measured data. The operational attribute of each work site is compared with all pre-defined operational attributes to match its corresponding hazard type value, denoted as H. j , where j is the index of any work site; to facilitate on-site ventilation location, several points are evenly set on the electronic layout map, where any one point is denoted as g, and g here refers to other points on the electronic layout map that do not include the work site; The distance d between each work site and point g is calculated using the Euclidean spatial distance formula. j (g), the specific formula is as follows: The three-dimensional coordinates of point g are (x(g), y(g), z(g)), and the three-dimensional coordinates of work site j are (x(g), y(g), z(g)). j y j , z j ); Take any point g and assign the hazard type value H to the work site. j and attenuation coefficient K j Through formula The risks at each point in the electronic layout diagram are quantified and superimposed to obtain the risk superposition value for each point. This is an exponentially decaying term, responsible for increasing the hazard type value H of the j-th work site. j According to distance d j (g) Attenuation; The risk overlay value of each point is marked on the electronic partial map to obtain the operation risk overlay map, and then transferred to step five. By introducing an exponential decay model based on physical laws, the propagation and superposition effects of risks of different natures in space are accurately quantified, generating a visualized operational risk superposition map. This transforms qualitative and local safety experience into a quantitative and global digital risk field, providing an intuitive and accurate spatial safety benchmark for global vehicle scheduling.

[0022] Step 5: Based on the operational risk overlay map, perform global optimization allocation of all vehicles to achieve collaborative safety management, specifically as follows: S5-1, Obtain the work route for each work site, identify the points on the work route and their corresponding risk superposition values, and calculate the average of the risk superposition values ​​of all points on the work route to obtain the route risk value, denoted as L. j Next, the risk type value H for the work site. j Its corresponding line risk value L j A weighting factor is assigned, and the risks faced during the operation at the work site are calculated by linear weighting and fusion to obtain the operation risk value. The operation risk value for each work site is denoted as G. j ; S5-2 establishes a negative correlation between vehicle condition and operational risk, i.e., the fit calculation formula; specifically, vehicles in good condition should be assigned to high-risk operations, and vehicles in poor condition should be assigned to low-risk tasks. The fit calculation formula is as follows: J rj This indicates the compatibility between vehicle r and work site j. The larger the value, the higher the matching degree. Qmin and Qmax refer to the minimum and maximum values ​​of the warning index among all vehicles that have not triggered the edge warning. Gmin and Gmax refer to the minimum and maximum values ​​of the work risk value among all work sites. Constraint settings: (1) Hard safety constraints: When the vehicle warning index q r Exceeding the warning safety threshold q safe At that time, it is prohibited to be assigned to high-risk work sites (i.e., G). j >G safe ), warning safety threshold q safe Primarily based on the statistical characteristics of historical operating data of the vehicle's three-electric system and functional safety degradation criteria, its principle lies in analyzing the distribution of vehicle warning indices under numerous normal and abnormal conditions to determine a critical point that can effectively distinguish between healthy operating conditions and conditions requiring restricted operation. Specifically, q safe The value is based on the high percentile (e.g., 90th percentile) of the historical vehicle warning index, with a certain safety margin added to ensure that when the vehicle's condition deteriorates to this threshold, its reliability has significantly decreased, and it must be restricted from operating in high-risk areas to prevent chain safety accidents caused by vehicle malfunctions in dangerous environments; High-risk warning G safe It is mainly based on site risk assessment standards and accident consequence severity classification. The principle is to comprehensively analyze the historical accident data, potential risk exposure frequency and possible personnel casualties and equipment damage of various work sites to determine an unacceptable risk level lower limit. This threshold usually corresponds to the work environment risk level that may cause major or higher-level accidents. (2) Operational capacity constraints: Ensure that the vehicle type matches the requirements of the operation task (e.g., excavators cannot be assigned to transportation tasks). Based on the fit degree calculation and constraint conditions, output the fit degree matrix J for all vehicle-work site combinations; S5-3, Global Optimization Matching Algorithm: Construct an optimization objective function with the goal of maximizing overall fit: , where X rj X is the decision variable, including 0 and 1. rj =1 indicates that vehicle r is assigned to work site j, X rj =0 indicates that vehicle r is not assigned to work location j; the following conditions must be met during the global optimization process: (4) At least one vehicle shall be assigned to each work site: ΣX rj ≥1, j; (5) Each vehicle can be assigned to a maximum of one work site: ΣX rj ≤1, r; (6) ΣX rj ≤W j , j(W) j (The maximum vehicle capacity of work site j). According to fit J rj Sort all possible vehicle-worksite pairs from high to low and assign them sequentially. If the constraints are met, confirm the assignment until all worksites have vehicles assigned or no vehicles can be assigned. Then output the optimal assignment scheme, that is, the worksite assignment result for each engineering vehicle. S5-4, Safety Collaborative Management Execution: The allocation results are sent to each vehicle terminal through the vehicle network, and risk warnings and precautions for the work site are sent at the same time. Then, real-time monitoring is carried out for dynamic adjustment: (1) When a vehicle completes the current work task, it enters the sequence of vehicles to be allocated; (2) When a vehicle has an edge warning, the vehicle's work is suspended, and the vehicle with the lowest warning index is selected from the vehicles to be allocated to continue to perform the work; In practical application, when the risk of a certain work site suddenly increases, all vehicles in the area are immediately notified, and vehicles with higher warning indices are given priority to evacuate or have their tasks adjusted. By coupling vehicle warning status with work site risks at the global level, a negatively correlated fit model is established, and global optimization is performed under multiple safety and physical constraints to assign the healthiest vehicles to the most dangerous tasks. This significantly improves the operational safety and system robustness of the engineering vehicle convoy without reducing operational efficiency, and achieves safe collaborative management of engineering vehicles.

[0023] This invention synchronously collects comprehensive data on the battery, motor, and electronic control systems of engineering vehicles. The data is then processed in real-time on the vehicle, extracting key features and calculating warning coefficients for each subsystem. If any coefficient exceeds a safety threshold, a local edge warning is immediately triggered; otherwise, the coefficient is increased and the next step is executed. For vehicles that do not trigger a warning, a warning index reflecting their overall health status is calculated through trend analysis based on their historical status data. Simultaneously, based on the electronic layout map of the construction site and risk attributes, a comprehensive operational risk overlay map of the entire site is generated through spatial overlay calculation. Finally, using the vehicle warning index and operational risk value as input, and aiming to maximize safety adaptability, global optimization matching is performed under various constraints to generate a vehicle task allocation scheme. This scheme is dynamically adjusted based on task completion status or sudden warnings to achieve collaborative safety management.

[0024] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A safety management method for the three-electric system of new energy engineering vehicles, comprising: Step one, synchronously collect data on the three key electrical components, including battery data, motor data, and electronic control data; characterized by, it also includes the following steps: Step 2: Based on the data of the three electrical components, perform edge feature extraction and analysis to obtain the warning coefficients of the battery, motor, and electronic control. If the warning coefficient of the battery, motor, or electronic control exceeds its corresponding threshold, an edge warning is triggered; otherwise, proceed to step 3. Step 3: For vehicles that have not triggered edge warnings, perform comprehensive state analysis based on all warning coefficients of their historical status to obtain the warning index; Step 4: Based on the electronic layout map of the construction site and the attributes of the work site, conduct site risk overlay analysis and output the work risk overlay map; Step 5: Based on the warning index and the operational risk overlay map, perform global optimization allocation for all vehicles that have not triggered edge warnings; specifically including: S5-1: Obtain the work route for each work site, identify the points on the work route and their corresponding risk superposition values, calculate the average of the risk superposition values ​​of all points on the work route, and obtain the route risk value, denoted as L. j Risk type value H for the work site j and the corresponding line risk value L j Assign a weighting factor and assign the risk type value H j and line risk value L j Perform linear weighted fusion and calculate the sum of risks faced in traveling to the work site to obtain the work risk value, denoted as G. j ; S5-2, Establish a negative correlation between vehicle status and operational risk, i.e., the fit calculation formula: J rj q represents the compatibility between vehicle r and work site j. r The warning index is the warning index for vehicles that have not triggered the edge warning. Qmin and Qmax refer to the minimum and maximum values ​​of the warning index among all vehicles that have not triggered the edge warning. Gmin and Gmax refer to the minimum and maximum values ​​of the operation risk value among all operation sites. Constraint settings: (1) Hard safety constraint: When the vehicle's warning index q exceeds the warning safety threshold q safe When the risk value G is high, vehicles are prohibited from being assigned to high-risk work sites; j Exceeding the operational risk threshold G safe (2) Operational capacity constraints: Ensure that the vehicle type matches the task requirements of the operation site; Based on the fit degree calculation and constraint conditions, output the fit degree matrix J for all vehicle-work site combinations; S5-3, Global Optimization Matching Algorithm: To maximize the overall fit, an optimization objective function is constructed: ； Among them, X rj For decision variables, including 0 and 1; X rj =1 indicates that vehicle r is assigned to work site j; X rj =0 indicates that vehicle r was not assigned to work site j; The following conditions must be met during the global optimization process: (1) At least one vehicle shall be assigned to each work site: ΣX rj ≥1, j; (2) Each vehicle can be assigned to a maximum of one work site: ΣX rj ≤1, r; (3) ΣX rj ≤W j , j, W j The maximum vehicle capacity of work site j; According to fit J rj Sort all possible vehicle-work site pairs from high to low and assign them sequentially. If the constraints are met, confirm the assignment until all work sites have vehicles assigned or no vehicles can be assigned. Output the optimal assignment scheme, that is, the assignment result of each engineering vehicle to the corresponding work site. S5-4, Safety Collaborative Management Execution: The allocation results are sent to each vehicle terminal via the vehicle network.

2. The safety management method for the three-electric system of new energy engineering vehicles according to claim 1, characterized in that, Step two specifically includes: S2-1, Edge feature extraction and analysis are performed based on battery data to obtain the battery early warning coefficient; S2-2, Edge feature extraction and analysis are performed based on motor data to obtain the motor early warning coefficient; S2-3, Edge feature extraction and analysis are performed based on electronic control data to obtain the electronic control early warning coefficient; S2-4: Preset battery warning threshold, motor warning threshold, and electronic control warning threshold respectively. Compare the battery warning coefficient, motor warning coefficient, and electronic control warning coefficient with the battery warning threshold, motor warning threshold, and electronic control warning threshold respectively. If the battery warning coefficient > the battery warning threshold, a battery warning is triggered; if the motor warning coefficient > the motor warning threshold, a motor warning is triggered; if the electronic control warning coefficient > the electronic control warning threshold, an electronic control warning is triggered. The edge warning includes battery warning, motor warning, and electronic control warning.

3. The safety management method for the three-electric system of new energy engineering vehicles according to claim 2, characterized in that, Step S2-1 specifically includes: The individual cell voltages of all cells in the battery pack are calculated using the standard deviation formula to determine the individual cell voltage consistency, thus obtaining the individual cell voltage coefficient. The temperature uniformity of all individual cells in the battery pack is calculated using the standard deviation formula, resulting in the individual cell temperature coefficient. Preset remaining battery power safety lower limit SOC safe If the remaining state of charge (SOC) of the battery is greater than or equal to the safe minimum SOC of the remaining charge... safe Then the remaining power coefficient is set to zero, i.e., R SOC =0; otherwise, the remaining power coefficient is: ;where clip(*) is the truncation function; The battery health status (SOH) is determined using a cutoff formula. The degradation state coefficient R was calculated. SOH ; Count the number of battery error reports between the last monitoring time and the current monitoring time, and use a truncation formula. Normalization is performed to obtain the battery error coefficient R. 电池 ; where m 电池 m is the number of times the battery has reported an error. ref This represents the maximum number of errors allowed within the monitoring period. A weighting factor is assigned to each of the cell voltage coefficient, cell temperature coefficient, remaining capacity coefficient, degradation state coefficient, and battery error coefficient. The cell voltage coefficient, cell temperature coefficient, remaining capacity coefficient, degradation state coefficient, and battery error coefficient are then fused together by linear weighting to obtain the battery warning coefficient.

4. A safety management method for the three-electric system of new energy engineering vehicles according to claim 3, characterized in that, Step S2-2 specifically includes: The motor rotor speed n is obtained through the formula Convert the angular velocity to ω, and then multiply the angular velocity by the actual output torque of the motor to obtain the mechanical power P. out Through engineering loss model The copper loss, iron loss, and mechanical loss were calculated separately; where k t The torque constant of the motor. This indicates the motor temperature T motor The phase resistance is shown below; a is the hysteresis loss coefficient; b is the eddy current loss coefficient; c is the mechanical loss coefficient. The heat generation rate Q is obtained by adding up the copper loss, iron loss, and mechanical loss. gen According to the thermal model The overheating risk coefficient λ is obtained; where, For the motor thermal resistance, For the heat capacity of the motor, T cool This refers to the coolant temperature. Count the number of motor error reports between the last monitoring time and the current monitoring time, and use a truncation formula. Normalization is performed to obtain the motor error coefficient R. 电机 ; where m 电机 m is the number of times the motor reported an error. ref This represents the maximum number of errors allowed within the monitoring period. A weighting factor is assigned to both the overheating risk coefficient and the motor error coefficient, and the two coefficients are then fused together using linear weighting to obtain the motor early warning coefficient.

5. A safety management method for the three-electric system of new energy engineering vehicles according to claim 4, characterized in that, Step S2-3 specifically includes: The temperature T of the electronically controlled controller contorl Using S-shaped response function Simulate a linear relationship and calculate the control risk coefficient β, T. critical is the critical temperature threshold of the controller chip in the electronic control system, which represents the highest operating temperature of the controller; h is the curve steepness factor, used to control the degree of drastic transition from low to high risk. The packet loss rate and latency rate during data transmission in the electronic control system are normalized. Specifically, the packet loss rate and latency rate are divided by the highest historical packet loss rate and latency rate in the electronic control system, respectively. Then, the normalized packet loss rate and latency rate are applied to a parallel failure model. The communication risk coefficient γ is calculated; where D is the packet loss rate and Y is the delay rate. Count the number of electronic control system errors between the last monitoring time and the current monitoring time, and use a truncation formula. Normalization is performed to obtain the electronic control error coefficient R. 电控 ; where m 电控 For the number of times the electronic control system reported an error; m ref This represents the maximum number of errors allowed within the monitoring period. A weighting factor is assigned to each of the control risk coefficient, communication risk coefficient, and electrical control error coefficient. The control risk coefficient, communication risk coefficient, and electrical control error coefficient are then fused together by linear weighting to obtain the electrical control early warning coefficient.

6. A safety management method for the three-electric system of new energy engineering vehicles according to claim 5, characterized in that, Step three specifically includes: The battery warning coefficient, motor warning coefficient, and electronic control warning coefficient of the vehicle are obtained from previous monitoring. A weight factor is assigned to each of the battery warning coefficient, motor warning coefficient, and electronic control warning coefficient. Then, the battery warning coefficient, motor warning coefficient, and electronic control warning coefficient are linearly weighted and fused to obtain the warning status value. A two-dimensional rectangular coordinate system is constructed with monitoring time as the horizontal axis and warning status value as the vertical axis. The warning status values ​​of each monitoring session are input into the coordinate system according to the corresponding monitoring time to form several discrete points, and then connected by a smooth curve to obtain the warning status value curve. The warning index for vehicles that have not triggered edge warnings is generated by smoothing the warning status value curve and mining the tangent slope trend, and then passed to step five.

7. A safety management method for the three-electric system of new energy engineering vehicles according to claim 6, characterized in that, The method for generating the warning index for vehicles that have not triggered edge warning specifically includes: Tangent lines are drawn to the curve at each discrete point in the warning status value curve, and the slope of the tangent lines is calculated using the least squares method. The average of all slopes greater than zero is used to calculate the increasing trend of the warning status value, denoted as f1. The absolute values ​​of the slopes less than zero are averaged again to calculate the decreasing trend of the warning status value, denoted as f2. The average of the warning status values ​​at each historical monitoring time is used to obtain the warning baseline value, denoted as F. The increasing trend, decreasing trend, and warning baseline value are then amplified using a trend amplification formula. The early warning index q is calculated.

8. A safety management method for the three-electric system of new energy engineering vehicles according to claim 7, characterized in that, In step four, the method for generating the operation risk overlay map includes: An electronic layout map of the entire construction site is pre-stored. This map marks various work areas requiring engineering vehicle operations and their associated work attributes. Each work attribute is pre-defined to correspond to a hazard type value and a reduction coefficient. The work attribute of each work area is compared with all pre-defined work attributes to match its corresponding hazard type value, denoted as H. j Where j is the index of any work site; several points are evenly set on the electronic layout map, where any one point is denoted as g, and g is other points on the electronic layout map that do not include the work site; The distance d between each work site and point g is calculated using the Euclidean spatial distance formula. j (g), the specific formula is as follows: Wherein, the three-dimensional coordinates of point g are (x(g), y(g), z(g)), and the three-dimensional coordinates of work site j are (x(g), y(g), z(g)) j y j , z j ); Take any point g and assign the hazard type value H to the work site. j and attenuation coefficient K j Through formula The risks at each point in the electronic layout map are quantified and superimposed to obtain the risk superposition value for each point. The risk superposition value of each point is marked on the electronic partial map to obtain the operation risk superposition map.

Images