Electric tricycle lithium battery charging and discharging management method and system

Abstract

The invention relates to the technical field of lithium battery safety, in particular to an electric tricycle lithium battery charging and discharging management method and system, and the method comprises the steps: obtaining lithium battery state data, environment data, tricycle load data and historical driving data; clustering analysis is carried out on the user driving data, and electro-tricycle scene types are divided; performing health degree evaluation on the battery state, automatically matching a charging mode according to the health degree, generating a charging current limiting value through a multi-objective optimization algorithm, performing dynamic matching on charging parameters, and executing charging operation according to the matched parameters; the scene type is recognized, a discharge current limiting value is generated through a multi-target optimization algorithm, discharge parameter dynamic matching is carried out, discharge operation is executed according to matched parameters, and energy recovery is started during downhill braking; and establishing a battery abnormal state identification model, and triggering graded early warning when an abnormal state is identified. Therefore, the problems of fixed charging and discharging modes, single abnormity monitoring, insufficient scene adaptation and the like in the prior art are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of lithium battery safety, and in particular relates to a lithium battery charging and discharging management method and system for an electric tricycle. BACKGROUND

[0002] In recent years, electric tricycles have been widely used in urban short-distance transportation, express delivery and rural walking due to their flexibility, convenience and strong carrying capacity. As the core power source of electric tricycles, the performance and life of lithium batteries directly affect the use cost and safety and reliability of the vehicle. However, the working environment of electric tricycles is complex and changeable, and often faces the alternating influence of different loads (empty load, full load), road conditions (flat road, slope, bumpy road section) and environmental conditions (high temperature, low temperature, humidity), which puts forward strict requirements on the charging and discharging management of lithium batteries.

[0003] At present, the traditional lithium battery charging and discharging management method has the following shortcomings: first, the charging mode mostly uses fixed parameter control, which cannot be dynamically adjusted according to the battery health state, and is easy to cause overcharging, undercharging or low charging efficiency of the battery, and long-term use will accelerate the battery decay; second, the discharging management is mostly based on simple voltage threshold control, which cannot be finely adjusted in combination with the actual driving scene, not only affecting the endurance capability, but also possibly causing safety hazards due to overloading discharge; third, the energy recovery mechanism lacks intelligent adaptation ability, and the recovery strength is fixed, which is difficult to balance the recovery efficiency maximization and driving safety under different road conditions; fourth, abnormal state monitoring mostly relies on single parameter threshold judgment, and the sensitivity of potential risk identification under complex scenes is insufficient, and the lagging early warning is easy to cause safety accidents.

[0004] In addition, the driving habits and use scenarios of electric tricycle users have significant differences, and the traditional "one-size-fits-all" charging and discharging strategy cannot meet the diversified needs. SUMMARY

[0005] The present application provides a lithium battery charging and discharging management method and system for an electric tricycle to solve the problems of fixed charging and discharging mode, single abnormal monitoring and insufficient scene adaptation in the prior art.

[0006] The first aspect embodiment of the application provides a lithium battery charging and discharging management method for an electric tricycle, including the following steps: obtaining lithium battery state data, environmental data, tricycle load data, and historical driving data; performing cluster analysis on the user driving data based on a reinforcement learning algorithm, dividing electric tricycle scene types, and establishing a scene feature library; dividing the charging process into four charging stages, each stage corresponding to different stage parameters; evaluating the health degree of the battery state according to the lithium battery state data, obtaining the health degree of the battery state, automatically matching the charging mode according to the health degree, generating a charging current limit value in real time based on the lithium battery state data and the charging mode through a multi-objective optimization algorithm, dynamically matching charging parameters based on the dynamically generated charging current limit value and the stage parameters, and performing a charging operation according to the matching parameters; identifying the scene type based on the scene feature library according to the environmental data and the tricycle load data, generating a discharging current limit value in real time based on the lithium battery state data and the scene type through a multi-objective optimization algorithm, dynamically matching discharging parameters based on the dynamically generated discharging current limit value, performing a discharging operation according to the matching parameters, starting energy recovery and matching the recovery intensity when braking on a downhill slope; real-time monitoring of lithium battery state, environmental data, and scene combinations, establishing a battery abnormal state identification model, triggering a hierarchical early warning when an abnormal state is identified, directly forcing the adjustment of the current limit value in a high-risk situation, and suspending dangerous operations.

[0007] Optionally, the lithium battery state data includes battery voltage, current, temperature, state of charge, and cycle count, and the environmental data includes environmental temperature, humidity, and altitude.

[0008] Optionally, the cluster analysis of the user driving data based on the reinforcement learning algorithm, the division of electric tricycle scene types, and the establishment of a scene feature library include: clustering historical driving data using an unsupervised clustering algorithm to generate multiple typical scene clusters; for each scene cluster, extracting a scene feature vector including average speed, driving mileage, load weight, and road slope; based on the Q-learning algorithm in reinforcement learning, optimizing and learning the scene feature vector to dynamically update and refine the scene feature library.

[0009] Optionally, the charging mode includes standard charging, fast charging, and maintenance charging, when the health degree is greater than a first target threshold, the fast charging mode is matched, otherwise, when the health degree is greater than a second target threshold, the standard charging mode is matched, otherwise, the maintenance charging mode is matched.

[0010] Optionally, the charging mode also supports custom selection.

[0011] Optionally, the four charging stages include a pre-charging stage, a constant current charging stage, a constant voltage charging stage and a trickle charging stage; and the stage parameters corresponding to the stages include a charging voltage threshold, a charging current threshold and a switching condition.

[0012] Optionally, the multi-objective optimization algorithm takes maximizing battery charging and discharging efficiency, maximizing battery cycle life and maximizing safety as optimization objectives.

[0013] Optionally, the battery abnormal state identification model is established by detecting voltage abnormality, current mutation and temperature overheating based on real-time monitored battery voltage, current and temperature data.

[0014] The second aspect embodiment of the application provides a lithium battery charging and discharging management system for an electric tricycle, comprising a data acquisition module, a scene modeling module, a charging module, a discharging module and a monitoring and safety protection module, wherein the data acquisition module is configured to acquire lithium battery state data, environment data, tricycle load data and historical driving data; the scene modeling module is configured to perform cluster analysis on the user driving data based on a reinforcement learning algorithm, divide electric tricycle scene types and establish a scene feature library; the charging module is configured to divide a charging process into four charging stages, each stage corresponding to different stage parameters, perform health degree evaluation on a battery state according to the lithium battery state data, obtain a health degree of the battery state, automatically match a charging mode according to the health degree, generate a charging current limit value in real time based on the lithium battery state data and the charging mode through a multi-objective optimization algorithm, perform charging parameter dynamic matching based on the dynamically generated charging current limit value and the stage parameters, and perform a charging operation according to the matching parameters; the discharging module is configured to identify a scene type based on the scene feature library according to the environment data and the tricycle load data, generate a discharging current limit value in real time according to the lithium battery state data and the scene type through a multi-objective optimization algorithm, perform discharging parameter dynamic matching based on the dynamically generated discharging current limit value, perform a discharging operation according to the matching parameters, start energy recovery and match recovery intensity when braking on a downhill slope; and the monitoring and safety protection module is configured to monitor lithium battery state, environment data and scene combinations in real time, establish a battery abnormal state identification model, trigger a hierarchical early warning when an abnormal state is identified, directly force adjustment of a current limit value in a high-risk situation, and suspend dangerous operations.

[0015] The application has the following advantages: The embodiment of the application can avoid damage operations such as overcharging and undercharging, effectively delay battery attenuation, and prolong service life by health degree evaluation and matching of adaptive charging modes combined with multi-stage dynamic charging parameter adjustment; based on battery state and scene feature dynamic generation of charge and discharge current limiting values, accurate parameter matching under different working conditions is realized, which can not only ensure the balance between charging speed and power sufficiency, but also efficiently release electric energy according to driving demand, and improve the endurance capability; the scene feature library is constructed through reinforcement learning, so that the discharge management can dynamically adjust the output strategy according to road conditions, load, environment and the like, intelligent energy recovery during downhill braking can further improve energy utilization rate, and meet the needs of diversified use scenarios; the abnormal state recognition model is established by fusing multi-dimensional data, the hierarchical early warning mechanism can respond to potential risks in time, and forced intervention under high-risk conditions can effectively avoid safety accidents such as battery thermal runaway and short circuit. Thus, the problems of fixed charging and discharging mode, single abnormal monitoring and insufficient scene adaptation in the prior art are solved.

[0016] Additional aspects and advantages of the application will be set forth in part in the description that follows, and in part will become apparent to those skilled in the art upon examination of the following description and drawings. BRIEF DESCRIPTION OF DRAWINGS

[0017] The above and / or additional aspects and advantages of the application will become apparent and be readily appreciated from the following description, including the accompanying drawings, in which: Figure 1 A flowchart of a lithium battery charging and discharging management method of an electric tricycle according to an embodiment of the application is provided. Figure 2 A structural schematic diagram of a lithium battery charging and discharging management system of an electric tricycle according to an embodiment of the application is provided. DETAILED DESCRIPTION

[0018] The embodiments of the application are described in detail below, examples of which are shown in the accompanying drawings, in which the same or similar reference numerals represent the same or similar elements or elements having the same or similar functions throughout. The embodiments described below are exemplary and are intended to explain the application, and cannot be understood as limiting the application.

[0019] An electric tricycle lithium battery charging and discharging management method and system are described below with reference to the accompanying drawings. In view of the problems of fixed charging and discharging mode, single abnormality monitoring, and insufficient scene adaptation mentioned in the background art, the present application provides an electric tricycle lithium battery charging and discharging management method and system. In this method, the health degree is evaluated to match and adapt the charging mode, and the multi-stage dynamic charging parameter adjustment is combined to avoid damaging operations such as overcharging and undercharging, effectively delay battery degradation, and prolong the service life. Based on the battery state and scene characteristics, the charging and discharging current limit value is dynamically generated to realize accurate parameter matching under different working conditions, which can not only balance the charging speed and power adequacy, but also efficiently release electric energy according to the driving demand to improve the endurance capability. The scene feature library is constructed through reinforcement learning, so that the discharge management can dynamically adjust the output strategy according to the road conditions, load, environment, etc. Intelligent energy recovery during downhill braking can further improve energy utilization and meet the needs of diversified use scenarios. The abnormal state recognition model is established by integrating multi-dimensional data, and the hierarchical early warning mechanism can respond to potential risks in time. Forced intervention in high-risk situations can effectively avoid safety accidents such as battery thermal runaway and short circuit. Thus, the problems of fixed charging and discharging mode, single abnormality monitoring, and insufficient scene adaptation in the prior art are solved.

[0020] Specifically, Figure 1 A flowchart of an electric tricycle lithium battery charging and discharging management method provided by the present application is shown.

[0021] As Figure 1 shown, the electric tricycle lithium battery charging and discharging management method includes the following steps: In step S101, the lithium battery state data, environmental data, tricycle load data, and historical driving data are obtained.

[0022] Among them, the single cell voltage is collected by the voltage sampling circuit integrated by the battery management system, and the charging and discharging current is detected by the Hall current sensor; NTC temperature sensors are embedded in the battery pack and distributed in the cell gap, and humidity and altitude sensors are installed externally; strain gauge pressure sensors are installed at the bottom of the cargo box to monitor the load; the GPS module records the driving trajectory, speed and altitude change, and the CAN bus connects the motor controller to obtain the output power and torque.

[0023] It can be understood that the embodiments of the application acquire lithium battery state data, environmental data, three-wheeled vehicle load data, and historical driving data, which can provide comprehensive and accurate basis for lithium battery charging and discharging management of the electric three-wheeled vehicle. Among them, the battery state data provides core parameters for health degree evaluation, charging mode matching, and charging and discharging current limit value generation, ensuring that the charging and discharging is adapted to the current state of the battery; the environmental data supports scene adaptation and ensures the stability of charging and discharging in different environments; the load data and the historical driving data are clustered and analyzed through the reinforcement learning algorithm, which helps to divide the scene types and establish a feature library, so that the discharge management can dynamically adapt to the road conditions and load, and optimize the energy recovery intensity; at the same time, the multi-dimensional data provides input for the abnormal state recognition model, improving the accuracy of safety monitoring and early warning.

[0024] In the embodiments of the application, the lithium battery state data includes battery voltage, current, temperature, state of charge, and cycle number, and the environmental data includes environmental temperature, humidity, and altitude.

[0025] Specifically, the voltage, current, temperature, state of charge, and cycle number are collected every 100 ms. The environmental temperature, humidity, and altitude are collected every 1 s.

[0026] It can be understood that the embodiments of the application collect the battery voltage, current, temperature, state of charge, and cycle number at a high frequency of 100 ms, and collect the environmental temperature, humidity, and altitude at a regular frequency of 1 s. The high-frequency data can accurately capture the instantaneous changes of the battery state (such as current mutation, voltage fluctuation, and temperature sudden rise), providing millisecond-level response basic data for health degree evaluation, charging and discharging current limit value dynamic generation, and real-time identification of abnormal state, ensuring the timeliness and accuracy of charging stage switching; and the regular frequency of collecting environmental parameters can meet the needs of scene recognition (such as altitude-related slopes and temperature and humidity affecting charging and discharging efficiency), while avoiding data redundancy and reducing system operation load, ultimately realizing accurate control of the lithium battery state and dynamic adaptation of the charging and discharging strategy, and improving management efficiency and safety.

[0027] In step S102, the user driving data is clustered and analyzed based on the reinforcement learning algorithm, the electric three-wheeled vehicle scene types are divided, and the scene feature library is established.

[0028] It can be understood that the embodiments of the application mine the internal rules of the data through the reinforcement learning algorithm, establish a scene feature library that can accurately map diversified use scenarios, and provide targeted support for the generation of current limit values in the discharge stage, energy recovery intensity matching, etc.

[0029] In the embodiments of the application, the user driving data is clustered and analyzed based on the reinforcement learning algorithm, the electric three-wheeled vehicle scene types are divided, and the scene feature library is established, including: The historical driving data is clustered by using an unsupervised clustering algorithm to generate a plurality of typical scenario clusters; For each scenario cluster, a scenario feature vector including average speed, driving mileage, load weight, and road slope is extracted; Based on the Q-learning algorithm in reinforcement learning, the scenario feature vector is optimized and learned to dynamically update and refine the scenario feature library.

[0030] Specifically, based on the Q-learning algorithm in reinforcement learning, the feature vector is continuously optimized and learned, taking the scenario recognition accuracy and the charging and discharging strategy adaptation effect as the reward signal, and through dynamic interaction with the actual driving environment, the feature weight is continuously adjusted to realize dynamic updating and refinement of the scenario feature library.

[0031] It can be understood that the embodiments of the present application cluster the historical driving data by using an unsupervised clustering algorithm, which can automatically aggregate typical scenario clusters with similar characteristics without pre-setting scenario labels, ensuring that the scenario division conforms to the actual use rules. For each scenario cluster, the core parameters are extracted to form a scenario feature vector, which abstracts the scenario into a quantifiable feature set, providing a clear basis for subsequent strategy adaptation. Then, through continuous optimization by the Q-learning algorithm, the feature library is dynamically updated. This makes the scenario division more in line with the actual situation, providing accurate basis for the discharging strategy, such as matching a larger current limit when climbing, adjusting the energy recovery intensity when descending, solving the problem of insufficient adaptation of the traditional "one-size-fits-all" mode, and adapting to different user driving habits, improving the intelligence and personalization level of lithium battery management, and optimizing the charging and discharging efficiency and battery life.

[0032] In step S103, the charging process is divided into four charging stages, each stage corresponding to different stage parameters.

[0033] In the embodiments of the present application, the four charging stages include a pre-charging stage, a constant current charging stage, a constant voltage charging stage, and a trickle charging stage. The stage parameters corresponding to each stage include a charging voltage threshold, a charging current threshold, and a switching condition.

[0034] It can be understood that the embodiments of the present application divide the charging process into four stages of pre-charging, constant current, constant voltage, and trickle charging and configure different parameters, which can realize fine management of lithium battery charging. In the pre-charging stage, a small current is used to activate the battery core to avoid large current impact. In the constant current stage, a larger current is used to quickly supplement the charge to improve the charging efficiency. In the constant voltage stage, the voltage is controlled to be constant to prevent overcharging while gradually filling the battery with electricity. In the trickle charging stage, a small current is used to finish charging to ensure that the battery is fully saturated. The stage parameters dynamically adapt to the battery state, solving the problems of overcharging and undercharging caused by traditional fixed parameter charging, balancing the charging speed and battery life, significantly reducing the battery degradation rate, and improving the charging safety and reliability.

[0035] In step S104, the health degree of the battery state is evaluated according to the lithium battery state data, the health degree of the battery state is obtained, the charging mode is automatically matched according to the health degree, the charging current limiting value is generated in real time based on the lithium battery state data and the charging mode through a multi-objective optimization algorithm, the charging parameters are dynamically matched based on the dynamically generated charging current limiting value and the stage parameters, and the charging operation is performed according to the matched parameters.

[0036] In the health degree evaluation of the battery state according to the lithium battery state data, the core indexes include the capacity attenuation rate, the cycle number and the internal resistance change, and the health degree is calculated by using a weighted algorithm. The capacity attenuation rate is the ratio of the current actual capacity to the initial capacity, the cycle number is the cumulative charging and discharging cycle number, and the internal resistance change is the difference between the current cell internal resistance and the initial internal resistance.

[0037] It can be understood that the capacity attenuation rate, the cycle number and the internal resistance change are used to comprehensively reflect the real state of the battery, the health degree of the battery is evaluated by using the weighted algorithm, the charging mode is associated according to the health degree of the battery, the battery is prevented from being damaged by improper charging, and the current limiting value generated by the multi-objective optimization can dynamically adapt the charging stage parameters, so that the safety is ensured, the charging efficiency and the battery life are balanced, the poor adaptability of the traditional fixed parameter charging is solved, and the charging is more intelligent, efficient and safe.

[0038] In the embodiment of the application, the charging mode includes standard charging, fast charging and maintenance charging. When the health degree is greater than a first target threshold, the fast charging mode is matched, otherwise, when the health degree is greater than a second target threshold, the standard charging mode is matched, otherwise, the maintenance charging mode is matched.

[0039] It can be understood that the charging mode is matched according to the health degree threshold in the embodiment of the application, so that the precise adaptation of the lithium battery charging can be realized. When the health degree is higher than the first target threshold, the fast charging is enabled to meet the demand of the user for the charging speed. When the health degree is between the first target threshold and the second target threshold, the standard charging is switched to balance between the efficiency and the battery protection. When the health degree is lower than the second target threshold, the maintenance charging is automatically entered to delay the attenuation through the equalization charging program. This dynamic matching mechanism not only avoids the damage of the fast charging to the battery when the health degree is low, but also improves the charging efficiency when the battery state is good, and prolongs the service life of the aging battery through the maintenance mode.

[0040] In the embodiment of the application, the charging mode also supports self-defined selection.

[0041] It can be understood that the embodiments of the present application can support custom selection on the basis of automatic matching of the system, further meeting the personalized needs of users. This flexibility not only retains the scientific nature of intelligent management, but also gives users the right to make independent decisions, balances the differences between system optimization and actual use scenarios, and improves user experience and operational convenience.

[0042] For example, when the battery health meets the fast charging condition but the user wants to reduce battery wear, the user can manually switch to standard charging; if the user faces an urgent travel demand, even if the battery health is in the standard charging interval, the user can still choose fast charging to shorten the waiting time; for users familiar with the characteristics of the battery, they can also adjust the parameters as needed outside the maintenance charging mode.

[0043] In the embodiments of the present application, the multi-objective optimization algorithm maximizes the battery charging and discharging efficiency, maximizes the battery cycle life, and maximizes the safety.

[0044] It can be understood that the embodiments of the present application adjust the charging and discharging parameters in real time, which not only improves the efficiency to shorten the charging time and increase the endurance, but also avoids sacrificing the battery life by excessively pursuing efficiency; by taking safety as the core constraint, the voltage, current, temperature and other parameters are strictly controlled within the safety threshold in the efficiency and life optimization, to prevent safety risks caused by overcharging, overdischarging or overheating. This multi-objective collaborative optimization mechanism solves the problem that efficiency, life and safety are difficult to balance in traditional single-target strategies, so that the lithium battery can work efficiently and run stably for a long time in different scenarios, and the use performance and reliability are improved.

[0045] In step S105, the scene type is identified based on the environment data and the three-wheeled vehicle load data according to the scene feature library, the discharging current limiting value is generated in real time based on the lithium battery state data and the scene type through the multi-objective optimization algorithm, the discharging parameter dynamic matching is performed based on the dynamically generated discharging current limiting value, the discharging operation is performed according to the matched parameters, the energy recovery is started when braking downhill and the recovery intensity is matched.

[0046] Among them, based on the scene feature library, the current scene type is identified according to the environment data and the three-wheeled vehicle load data through the cosine similarity algorithm. The cosine similarity algorithm calculates the cosine value of the angle between two vectors to measure the similarity between the real-time collected data and the typical scene features in the scene feature library. If the similarity is greater than the first similarity threshold, the corresponding scene is directly matched, otherwise, if the similarity is greater than the second similarity threshold, the historical driving trajectory is combined to assist in determining, otherwise, it is marked as a "new scene" and the feature library updating mechanism is triggered.

[0047] The energy recovery during downhill braking needs to meet multi-dimensional triggering conditions, including slope conditions and brake signals and battery state constraints. When the slope is greater than the target slope threshold, the energy recovery preparation state is activated; when the brake pedal stroke is greater than the target threshold or the hand brake is pulled up, the energy recovery is formally started. The recovery needs to ensure that the state of charge is less than the target power threshold and there is no battery failure.

[0048] When the starting condition is met, the motor controller receives the recovery intensity instruction, switches to the power generation mode, and recovers the current after rectification and filtering Input the battery pack, while uploading the recovery current and voltage data in real time through the CAN bus.

[0049] It can be understood that the embodiments of the application identify the scene type based on the scene feature library, generate the discharge current limiting value through multi-objective optimization in combination with the battery state, and dynamically match the parameters, while adapting the energy recovery intensity during downhill braking. It can not only make the discharge strategy accurately fit the power demand of different scenes such as heavy load climbing and empty load flat road, balance output and battery load through dynamic current limiting, but also maximize the use of downhill kinetic energy by means of energy recovery, and improve the endurance.

[0050] In step S106, the lithium battery state, environmental data and scene combination are monitored in real time, a battery abnormal state identification model is established, and a hierarchical early warning is triggered when an abnormal state is identified. In the case of high risk, the current limiting value is forcibly adjusted and the dangerous operation is stopped.

[0051] The hierarchical early warning includes low-risk early warning, medium-risk early warning and high-risk early warning. When the low-risk early warning is triggered, the abnormal type is announced by voice; when the medium-risk early warning is triggered, the charging and discharging power is automatically reduced to 50% to 70% of the rated value, and the vehicle warning light flashes; when the high-risk early warning is triggered, the charging and discharging circuit is immediately cut off, and the sound and light alarm is triggered.

[0052] It can be understood that the embodiments of the application monitor the lithium battery state, environmental data and scene combination in real time. Multi-dimensional data fusion makes abnormal identification more accurate, avoids false positives or false negatives of single parameter monitoring; the hierarchical early warning triggered by the abnormal state identification model can take measures such as prompting and reducing power according to the risk level, balance safety and use experience; the forced current limiting and operation stop in the case of high risk can directly block the danger such as over-temperature and short circuit, prevent safety accidents from the source, and solve the problems of traditional monitoring lag and single response. It comprehensively safeguards the safe operation of lithium batteries and vehicles.

[0053] In the embodiments of the application, the battery abnormal state identification model is established, including: Based on the real-time monitored battery voltage, current and temperature data, the isolated forest algorithm is used to detect voltage abnormalities, current surges and temperature overheating.

[0054] It should be noted that the isolation forest is an efficient unsupervised anomaly detection algorithm, which models the characteristic that "abnormal samples are more likely to be isolated" by building multiple isolation trees. For each sample, an isolation tree is constructed: a feature is randomly selected, a split point is randomly selected within the value range of the feature, the sample is recursively divided until each leaf node contains only one sample. Abnormal samples deviate from the normal range of features and are divided earlier.

[0055] For the real-time input feature vector (12 features of the above voltage, current, temperature), the average path length in all isolation trees is calculated, which is converted into an anomaly score. The closer the score is to 1, the more likely the sample is abnormal.

[0056] It can be understood that the embodiments of the present application detect voltage anomalies, current surges and temperature overheating based on real-time monitored voltage, current and temperature data using the isolation forest algorithm. Without a large number of labeled samples, the application can accurately identify abnormalities, acutely capture subtle changes such as voltage drop, current surge and temperature rise, avoid the rule rigidity and false alarm and missed alarm problems of traditional threshold method; multi-dimensional data fusion makes the abnormal judgment more comprehensive, which can identify short circuit, thermal runaway and other risks in advance, provides reliable basis for grading early warning, reduces the probability of safety accidents from the source, and improves the intelligence and safety of lithium battery monitoring.

[0057] According to the electric tricycle lithium battery charging and discharging management method provided by the embodiments of the present application, the health degree is evaluated to match the adaptive charging mode, and the multi-stage dynamic charging parameter adjustment is combined, so that the damaging operation such as overcharging and undercharging can be avoided, the battery attenuation can be effectively delayed, and the service life can be prolonged; based on the battery state and scene characteristics, the charging and discharging current limiting value is dynamically generated, the parameter accurate matching under different working conditions is realized, the balance between charging speed and power sufficiency can be ensured, the electric energy can be efficiently released according to the driving demand, and the endurance capability is improved; the scene characteristic library is constructed through reinforcement learning, so that the discharge management can dynamically adjust the output strategy according to the road conditions, load, environment and the like, the intelligent energy recovery during downhill braking can further improve the energy utilization rate, and the diversified use scene demand can be met; the multi-dimensional data is fused to establish an abnormal state identification model, the grading early warning mechanism can respond to potential risks in time, and the forced intervention under high-risk conditions can effectively avoid safety accidents such as battery thermal runaway and short circuit. Thus, the problems of fixed charging and discharging mode, single abnormal monitoring and insufficient scene adaptation in the prior art are solved.

[0058] Secondly, a kind of electric tricycle lithium battery charging and discharging management system according to the embodiments of the present application is described with reference to the drawings.

[0059] Figure 2 It is a structure schematic diagram of the electric tricycle lithium battery charging and discharging management system of the embodiments of the present application.

[0060] As Figure 2As shown, the lithium battery charging and discharging management system 10 of the electric tricycle includes a data acquisition module 100, a scene modeling module 200, a charging module 300, a discharging module 400, and a monitoring and safety protection module 500.

[0061] The data acquisition module 100 is configured to acquire lithium battery state data, environmental data, tricycle load data, and historical driving data. The scene modeling module 200 is configured to perform clustering analysis on user driving data based on a reinforcement learning algorithm, divide electric tricycle scene types, and establish a scene feature library. The charging module 300 is configured to divide the charging process into four charging stages, each stage corresponding to different stage parameters. The health of the battery state is evaluated based on the lithium battery state data, and the health of the battery state is obtained. The charging mode is automatically matched based on the health. Based on the lithium battery state data and the charging mode, the charging current limit value is generated in real time through a multi-objective optimization algorithm. The charging parameter is dynamically matched based on the dynamically generated charging current limit value and the stage parameter. The charging operation is performed according to the matched parameters. The discharging module 400 is configured to identify the scene type based on the scene feature library, the environmental data, and the tricycle load data. The discharging current limit value is generated in real time through a multi-objective optimization algorithm based on the lithium battery state data and the scene type. The discharging parameter is dynamically matched based on the dynamically generated discharging current limit value. The discharging operation is performed according to the matched parameters. The output is adjusted according to the road conditions and the load during discharging. Energy recovery is started and the recovery intensity is matched when braking on a downhill slope. The monitoring and safety protection module 500 is configured to monitor the lithium battery state, environmental data, and scene combination in real time, establish a battery abnormal state recognition model, trigger a hierarchical early warning when an abnormal state is recognized, directly force the adjustment of the current limit value in high-risk situations, and stop dangerous operations.

[0062] It should be noted that the foregoing explanation and description of the electric tricycle lithium battery charging and discharging management method embodiment also applies to the electric tricycle lithium battery charging and discharging management system of the embodiment, which will not be described here.

[0063] According to the electric tricycle lithium battery charging and discharging management system provided by the embodiment of the present application, the health degree is evaluated to match the adaptive charging mode, and the multi-stage dynamic charging parameter adjustment can avoid damage operations such as overcharging and undercharging, effectively delay battery attenuation, and prolong the service life; the charging and discharging current limiting value is dynamically generated based on the battery state and scene characteristics, the parameter accurate matching under different working conditions is realized, the balance between the charging speed and the power sufficiency can be ensured, the electric energy can be efficiently released according to the driving demand, and the endurance capability is improved; the scene characteristic library is constructed through reinforcement learning, the discharge management can dynamically adjust the output strategy according to the road conditions, load, environment and the like, the intelligent energy recovery during downhill braking can further improve the energy utilization rate, and the diversified use scene demand is met; the abnormal state recognition model is established by fusing multi-dimensional data, the grading early warning mechanism can timely respond to potential risks, and the forced intervention under the high-risk condition can effectively avoid safety accidents such as battery thermal runaway and short circuit. Thus, the problems of fixed charging and discharging mode, single abnormal monitoring and insufficient scene adaptation in the prior art are solved.

[0064] In the description of the present specification, the description of the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" and the like means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or N embodiments or examples in a suitable manner. In addition, the person skilled in the art can combine and combine the different embodiments or examples described in the present specification and the features of the different embodiments or examples without contradiction.

[0065] In addition, the terms "first", "second" are only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features limited by "first", "second" can explicitly or implicitly include at least one of the features. In the description of the present application, the meaning of "N" is at least two, such as two, three, etc., unless otherwise specifically limited.

[0066] Any process or method descriptions in flow charts or otherwise described herein can be understood as representing code modules, segments, or portions of code that include one or more executable instructions for implementing specific logic functions (or steps) in the process, and the preferred embodiments of the present application include additional implementations in which the order of the steps can be different, including use of the same step for different functions, and the functions can be performed in the same or in other orders, including substantially concurrently or in reverse order, depending upon the functionality involved. The description herein of the preferred embodiments of the application is not intended to be exhaustive or to be necessarily limited to the precise steps or construction described. The steps and constructions may vary depending upon the implementation chosen and can be implemented in several different ways.

[0067] It should be understood that portions of the application can be implemented in hardware, software, firmware, or combinations thereof. In the above embodiments, the N steps or methods can be implemented in software or firmware stored in a memory and executed by a suitable instruction execution system. As such, if implemented in hardware, and in another embodiment, any of the following technologies, known in the art, or their combinations can be used: discrete logic circuitry having logic gates for implementing logic functions on data signals, application specific integrated circuits having appropriate combinational logic gates, programmable gate arrays (PGA), field programmable gate arrays (FPGA), and the like.

[0068] Those skilled in the art of the present technology can understand that all or part of the steps carried out by the above-mentioned embodiment methods can be completed by programs instructing related hardware, and the programs can be stored in a computer readable storage medium. When the program is executed, it includes one of the steps of the method embodiment or a combination thereof.

Claims

1. A method for charging and discharging management of a lithium battery of an electric tricycle, characterized in that, The method comprises the following steps: Obtain lithium battery state data, environmental data, tricycle load data, historical driving data; Based on the reinforcement learning algorithm, the user driving data is clustered and analyzed, the electric tricycle scene type is divided, and the scene feature library is established; The charging process is divided into four charging stages, each stage corresponding to different stage parameters; According to the lithium battery state data, the health degree of the battery state is evaluated, the health degree of the battery state is obtained, the charging mode is automatically matched according to the health degree, and the charging current limiting value is generated in real time based on the lithium battery state data and the charging mode through a multi-objective optimization algorithm. Based on the dynamically generated charging current limiting value and the stage parameters, the charging parameters are dynamically matched, and the charging operation is performed according to the matched parameters; Based on the scene feature library, the scene type is identified according to the environmental data and the tricycle load data, the discharge current limiting value is generated in real time according to the lithium battery state data and the scene type through a multi-objective optimization algorithm, the discharge parameters are dynamically matched based on the dynamically generated discharge current limiting value, and the discharge operation is performed according to the matched parameters. When braking downhill, start energy recovery and match the recovery intensity; Real-time monitoring of lithium battery state, environmental data and scene combination, establishment of battery abnormal state identification model, triggering of hierarchical early warning when abnormal state is identified, direct forced adjustment of current limiting value in high-risk situation, and termination of dangerous operation.

2. The electric tricycle lithium battery charging and discharging management method according to claim 1, characterized in that, The lithium battery state data includes battery voltage, current, temperature, state of charge and cycle number, and the environmental data includes environmental temperature, humidity and altitude.

3. The electric tricycle lithium battery charging and discharging management method according to claim 1, characterized in that, The clustering analysis of the user driving data based on the reinforcement learning algorithm, the division of the electric tricycle scene type, and the establishment of the scene feature library comprise: Using unsupervised clustering algorithm to cluster historical driving data to generate multiple typical scene clusters; For each scene cluster, extract the scene feature vector including average speed, driving mileage, load weight and road slope; Based on the Q-learning algorithm in reinforcement learning, the scene feature vector is optimized and learned to dynamically update and refine the scene feature library.

4. The electric tricycle lithium battery charging and discharging management method according to claim 1, characterized in that, The charging mode includes standard charging, fast charging and maintenance charging. When the health degree is greater than a first target threshold, the fast charging mode is matched, otherwise, when the health degree is greater than a second target threshold, the standard charging mode is matched, otherwise, the maintenance charging mode is matched.

5. The electric tricycle lithium battery charge and discharge management method according to claim 4, characterized in that, The charging mode also supports custom selection.

6. The electric tricycle lithium battery charge and discharge management method according to claim 1, characterized in that, The four charging stages include pre-charging stage, constant current charging stage, constant voltage charging stage and trickle charging stage. The stage parameters corresponding to each stage include charging voltage threshold, charging current threshold and switching condition.

7. The electric tricycle lithium battery charge and discharge management method according to claim 1, characterized in that, The multi-objective optimization algorithm maximizes battery charging and discharging efficiency, maximizes battery cycle life, and maximizes safety.

8. The electric tricycle lithium battery charge and discharge management method according to claim 1, characterized in that, The establishment of the battery abnormal state identification model comprises: Based on the real-time monitored battery voltage, current and temperature data, the isolated forest algorithm is used to detect voltage abnormalities, current surges and temperature overheating.

9. A three-wheeled electric vehicle lithium battery charge and discharge management system characterized by, It comprises: Data acquisition module, scene modeling module, charging module, discharging module, monitoring and safety protection module, wherein, The data acquisition module is configured to acquire lithium battery state data, environment data, tricycle load data, and historical driving data; The scene modeling module is configured to perform cluster analysis on the user driving data based on a reinforcement learning algorithm, divide electric tricycle scene types, and establish a scene feature library; The charging module is configured to divide a charging process into four charging stages, each stage corresponding to different stage parameters, perform health degree evaluation on the battery state according to the lithium battery state data, obtain the health degree of the battery state, automatically match a charging mode according to the health degree, generate a charging current limit value in real time based on the lithium battery state data and the charging mode through a multi-objective optimization algorithm, perform charging parameter dynamic matching based on the dynamically generated charging current limit value and the stage parameters, and perform a charging operation according to the matching parameters; The discharging module is configured to identify a scene type based on the scene feature library according to the environment data and the tricycle load data, generate a discharging current limit value in real time based on the lithium battery state data and the scene type through a multi-objective optimization algorithm, perform discharging parameter dynamic matching based on the dynamically generated discharging current limit value, perform a discharging operation according to the matching parameters, and start energy recovery and match the recovery intensity when braking on a downhill slope; The monitoring and safety protection module is configured to monitor the lithium battery state, environment data, and scene combination in real time, establish a battery abnormal state recognition model, trigger a hierarchical early warning when an abnormal state is recognized, directly force adjustment of the current limit value in a high-risk situation, and suspend dangerous operations.

Images