New energy automobile intelligent DCDC converter integrated charger

Abstract

The invention provides a new energy automobile intelligent DCDC converter integrated charger, and relates to the technical field of new energy automobile power electronics, the charger comprises a main circuit module, a sensor module, a data processing module, a machine learning module, a control execution module and a communication module; the main circuit module is integrated with a vehicle-mounted charger, a DCDC conversion unit and a high-voltage power distribution unit, so that equipment miniaturization is realized; the sensor module collects multi-dimensional parameters, the multi-dimensional parameters are analyzed by the machine learning module after data processing, a built-in prediction model can predict the service life and the fault risk of a part, and a dynamic adjustment model generates a power adjustment strategy by combining road conditions and battery states; and the control execution module realizes power regulation, grading protection and mode smooth switching according to the analysis result. Through integrated design and an intelligent algorithm, the energy conversion efficiency, the system safety and the working condition adaptability are improved, and the efficient, safe and intelligent requirements of the new energy automobile are met.

Description

Technical Field

[0001] The present invention relates to the field of power electronics technology for new energy vehicles, and in particular to an intelligent DC-DC converter-integrated charger for new energy vehicles. Background Art

[0002] Electric vehicles, hybrid vehicles, and other vehicles are equipped with various power conversion devices, including inverters for driving motors, chargers for charging high-voltage batteries from commercial power sources, and DC-DC converters for supplying power to auxiliary batteries. These chargers and power conversion devices use switching circuits that generate significant high-frequency noise.

[0003] The rapid development of the new energy vehicle industry is placing higher demands on the performance, reliability, and integration of on-board charging equipment. Traditional on-board chargers and DC-DC converters are independent of each other, resulting in large size, heavy weight, high cost, and low energy conversion efficiency. Furthermore, precise power regulation and intelligent management are difficult to achieve under complex driving conditions and fluctuating battery states, failing to fully meet the development needs of efficient, safe, and intelligent new energy vehicles. For example, under different road conditions, the charging and discharging requirements of vehicles vary significantly, and traditional chargers are unable to respond quickly and optimize energy distribution. They also lack effective protection and adaptive regulation mechanisms for battery aging or temperature abnormalities.

[0004] Therefore, it is necessary to provide an integrated charger with intelligent DCDC converter for new energy vehicles to solve the above technical problems. Summary of the Invention

[0005] The present invention aims to provide a charger with an integrated intelligent DC-DC converter for new energy vehicles. Through a highly integrated design, intelligent control algorithms, and a multi-mode collaborative working mechanism, the charger can be miniaturized, lightweight, efficient, and intelligent, thereby improving the overall performance and safety of new energy vehicles.

[0006] To solve the above technical problems, the present invention provides a new energy vehicle intelligent DCDC converter integrated charger, comprising: The main circuit module integrates the on-board charger unit, DCDC conversion unit and high-voltage power distribution unit; The sensor module includes a plurality of sensor groups, which are respectively arranged on each key component of the main circuit module and are used to collect electrical parameters and physical state parameters of the corresponding components; A data processing module connected to the sensor module, configured to receive and pre-process the electrical parameters and physical state parameters; wherein the pre-processing includes filtering, calibration, normalization, and feature extraction; A machine learning module connected with the data processing module, which is internally provided with a trained prediction model and a dynamic adjustment model, is used to analyze and calculate the preprocessed electrical parameters and physical state parameters; A control execution module connected with the machine learning module and the main circuit module is used to control the operation state of the main circuit module according to the output of the machine learning module; A communication module is used to realize data interaction with the vehicle controller and the battery management system, and receive road condition information and battery state information.

[0007] The sensor module according to the preferred technical solution of the present application comprises: A voltage sensor group is used to collect input voltage, output voltage and voltage of each key node; A current sensor group is used to collect input current, output current and current of each branch; A temperature sensor group is used to collect temperature of power devices, temperature of inductors and ambient temperature; A vibration sensor is used to collect vibration parameters of key components; An insulation monitoring sensor is used to monitor the insulation state of the system.

[0008] The machine learning module according to the preferred technical solution of the present application comprises: A predictive maintenance unit is used to predict the remaining useful life and potential failure risk of each component through the prediction model based on the preprocessed electrical parameters and physical state parameters; A dynamic energy distribution unit is used to generate a DCDC output power adjustment strategy through the dynamic adjustment model in combination with the road condition information and battery state information; A model updating unit is used to perform online optimization on the prediction model and dynamic adjustment model according to actual operation data.

[0009] The predictive maintenance unit according to the preferred technical solution of the present application comprises: An input feature processing subunit is used to perform standardization processing on the preprocessed time series parameters; An LSTM prediction network is used to process time series features through an improved gating mechanism to realize time series analysis on the component state; A remaining useful life calculation subunit is used to calculate the predicted remaining useful life based on the component degradation amount evolution curve and failure threshold; A failure risk assessment subunit is used to calculate the health degree based on the degradation amount and match a five-level failure warning level to output a failure warning level; A model optimization subunit is used to optimize the prediction model parameters by using an Adam optimizer and a joint loss function.

[0010] As a preferred technical solution of the present invention, the dynamic adjustment model includes a road condition identification sub-model and a power optimization sub-model: The road condition identification sub-model identifies the current road type, slope and congestion status based on the received road condition information; The power optimization sub-model dynamically adjusts the output voltage and current of the DCDC converter according to the road condition recognition result and the battery status information.

[0011] The control execution module includes: A power regulation unit, used to regulate the output power of the DCDC converter; a protection control unit, executing overvoltage, overcurrent, and overheat protection actions based on the output of the predictive maintenance unit and real-time electrical parameters and physical state parameters; The mode switching unit is used to realize automatic switching of charging and discharging modes and power levels.

[0012] As a preferred technical solution of the present invention, the present invention further includes a storage module for storing historical electrical parameters and physical states, model parameters and fault records.

[0013] As a preferred technical solution of the present invention, the protection control unit performs hierarchical protection actions in combination with the fault warning level: When the fault warning level is level 2, the output power is reduced to the rated value of the preset level 1 ratio, and the cooling system is controlled to run at the preset low speed; When the fault warning level is level three, the output power is reduced to the rated value of the preset level two ratio, and the cooling system is controlled to run at the preset high speed; When the fault warning level is level 4, the preset non-critical load output is turned off; When the fault warning level reaches level five, the main circuit relay is cut off, triggering the vehicle fault alarm and uploading data to the cloud.

[0014] Compared with related technologies, the new energy vehicle intelligent DCDC converter integrated charger provided by the present invention has the following beneficial effects: 1. This invention integrates an onboard charger, DC-DC converter, and high-voltage power distribution unit, reducing the redundant structure of traditional independent equipment and lowering energy transmission losses. Furthermore, a dynamic adjustment model optimizes power distribution based on road conditions and battery status, further improving energy efficiency and achieving device miniaturization and lightweighting. 2. This invention uses multiple sensors to collect electrical and physical parameters in real time, combines it with an LSTM network to predict the remaining life and failure risk of components, and implements overvoltage, overcurrent, and overheating protection through a hierarchical protection mechanism to provide early warning of potential faults, avoid sudden failures, and ensure the stable operation of the vehicle's power system.

[0015] 3. The application accurately judges the road type, slope and congestion state through the road condition recognition sub-model, dynamically adjusts the output power by linking the battery health and temperature parameters, and matches the different driving working condition requirements; the linear transition algorithm is adopted during mode switching to avoid voltage and current impact, and the user experience and equipment adaptability are improved.

[0016] In summary, the application realizes the miniaturization of the device by integrating multiple modules to reduce energy loss, predicts faults and performs hierarchical protection using multiple sensors and intelligent algorithms to improve safety, dynamically adjusts power in combination with road conditions and battery state and optimizes mode switching through a linear transition algorithm, achieving high efficiency, high reliability and intelligent adaptation of new energy vehicle charging equipment. BRIEF DESCRIPTION OF DRAWINGS

[0017] Fig. 1 The connection relationship diagram of the overall module of the new energy vehicle intelligent DCDC converter integrated charger provided by the application is shown in the following figure: Fig. 2 The internal unit connection relationship diagram of the machine learning module provided by the application is shown in the following figure. DETAILED DESCRIPTION

[0018] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the application.

[0019] The terms used in the present disclosure are only for the purpose of describing specific embodiments, and are not intended to limit the present disclosure. The singular forms "a", "an" and "the" used in the present disclosure and the appended claims are also intended to include the plural forms, unless the context clearly indicates otherwise. It should also be understood that the term "and / or" used herein means and includes any or all possible combinations of one or more associated listed items.

[0020] It should be understood that although the terms first, second, third, etc. may be used in the present disclosure to describe various information, these information should not be limited to these terms. These terms are only used to distinguish information of the same type from each other. For example, without departing from the scope of the present disclosure, the first information can also be referred to as the second information, and similarly, the second information can also be referred to as the first information. Depending on the context, the word "if" as used herein can be interpreted as "when" or "upon determination" or "in response to determining".

[0021] Please refer to Figs. 1-2 . The new energy vehicle intelligent DCDC converter integrated charger comprises: The main circuit module integrates an on-board charger unit, a DC-DC converter unit, and a high-voltage power distribution unit; the on-board charger unit is used to convert external AC power into DC power, the DC-DC converter unit is used to convert high-voltage DC power into low-voltage DC power, and the high-voltage power distribution unit is used to distribute and protect the high-voltage power supply; The sensor module includes multiple sensor groups, which are respectively arranged on each key component of the main circuit module to collect electrical parameters and physical state parameters of the corresponding components; wherein the key components include but are not limited to the power devices, inductors, capacitors, input and output ports and housing of the main circuit module; The data processing module is connected to the sensor module via a signal cable and is used to receive and pre-process the original electrical parameters and physical state parameters; the pre-processing includes filtering, calibration, normalization and feature extraction; The machine learning module is connected to the data processing module and has a built-in trained prediction model and dynamic adjustment model for analyzing and calculating pre-processed electrical parameters and physical state parameters; A control execution module is connected to the machine learning module and the main circuit module through a control bus, and is used to receive the control instructions output by the machine learning module and convert them into drive signals to control the operating state of the main circuit module; The communication module is connected to the vehicle controller and battery management system through the on-board network interface. It is used to realize data interaction with external systems and receive road condition information including road slope, vehicle speed, and traffic conditions, as well as battery status information including battery charge status, health status, and temperature.

[0022] It should be noted that the composition and functions of each core module in the charger and the integrated design of the main circuit module reduce the redundancy of traditional independent equipment. The multi-dimensional acquisition of the sensor module provides a basis for intelligent analysis. Each module forms a closed loop of acquisition, processing, analysis and control through data interaction, which not only realizes the miniaturization of the equipment, but also lays the hardware and data foundation for subsequent intelligent adjustment and protection, and improves the system integration and collaborative efficiency.

[0023] In the present invention, the sensor module includes: The voltage sensor group includes an AC voltage sensor at the AC input end, a DC voltage sensor at the DC output end, and node voltage sensors at key nodes of the main circuit, which are used to collect input voltage. , high voltage DC output voltage , low voltage DC output voltage And the node voltage of each key node (n=1,2,...,N, N is the number of nodes), and recorded as the voltage feature subset ; The current sensor group includes current sensors set in the input main circuit, output main circuit and each branch circuit to collect the total AC input current. , total DC input current , DC output total current and branch current (m=1,2,...,M, M is the number of branches), and recorded as the current feature subset ; The temperature sensor group includes a contact sensor attached to the surface of the power device, an inductor temperature sensor embedded in the inductor winding, a capacitor temperature sensor attached to the capacitor housing, and an ambient temperature sensor, which is used to collect the temperature of the power device. (k=1,2,...,K, K is the number of power devices), inductor temperature , capacitor temperature and ambient temperature , and recorded as the temperature feature subset ; Vibration sensor, used to collect vibration parameters of key components: MEMS triaxial acceleration sensor is installed in the main circuit module housing and key component installation position to collect X, Y, and Z axis vibration acceleration , and then calculate the effective value of vibration , the calculation formula is: , Indicates the integral within the set time window; the X, Y, and Z axis vibration acceleration and vibration effective value Recorded as vibration feature subset ; Insulation monitoring sensor, used to monitor the insulation status of the system: the balanced bridge method is used to monitor the insulation status of the system, including a first measurement branch connected in series between the high-voltage positive pole and the ground and a second measurement branch connected in series between the high-voltage negative pole and the ground. Each measurement branch contains a reference resistor of known resistance and a controllable switch; by controlling the on and off state of the switch, the voltage of the measurement point under different states is collected, and then the insulation resistance is calculated. , which calculates the insulation resistance Calculated by the following formula: ,in is the resistance of the known reference resistor, is the voltage across the reference resistor, is the voltage value at the measuring point, ) is the high voltage DC bus voltage; the insulation resistance of each measuring point is It is denoted as the insulation resistance feature subset R.

[0024] It should be noted that by refining the layout position, collection parameters and data processing methods of each sensor group, comprehensive monitoring of the main circuit electrical performance (voltage, current), component status (temperature, vibration) and system safety (insulation) is achieved. The multi-dimensional feature subset provides precise input for the state prediction and dynamic adjustment of the subsequent machine learning module, which not only ensures the integrity of monitoring, but also lays a data foundation for intelligent control, and improves the accuracy and comprehensiveness of system status perception.

[0025] In the present invention, the machine learning module includes: Predictive maintenance unit, which uses prediction models to predict the remaining service life and potential failure risks of each component based on electrical parameters and physical status parameters; The dynamic energy allocation unit combines road condition information and battery status information to generate a DCDC output power adjustment strategy through a dynamic adjustment model; The model update unit is used to perform online optimization of the prediction model and dynamic adjustment model based on actual operation data. The parameter update formula is: ,in is the updated parameter, is the parameter before updating, is the learning rate, is the parameter of the i-th local model, and N is the number of models involved in the aggregation.

[0026] It should be noted that by setting up a machine learning module, the predictive maintenance unit can predict component status, the dynamic energy allocation unit can optimize power regulation, and the model update unit can support online iteration of the algorithm. The collaboration of the three not only ensures the system's proactive maintenance and efficient energy management, but also adapts to different working conditions through model self-optimization, thereby improving the intelligence level and long-term applicability of the solution.

[0027] In the present invention, the predictive maintenance unit includes: The input feature processing subunit receives the time series parameters of the electrical parameters and physical state parameters preprocessed by the data processing module. The time series parameters include the voltage feature subset, the current feature subset, the temperature feature subset, the vibration feature subset, and the insulation resistance feature subset, and normalizes each feature subset to the interval [0-1]. The LSTM prediction network consists of an input layer, three hidden layers, and two output layers. The input layer has a 16-dimensional dimension (corresponding to 16 feature parameters), and each hidden layer contains 64 LSTM units. The hidden layer processes time series features by improving the gating mechanism of the LSTM unit. The core calculation includes the input gate, forget gate, output gate, and cell state update, which can be expressed as follows: The forget gate is used to determine the retention ratio of the cell state at the previous moment ,in is the output of the forget gate at time t (a value between 0 and 1, the closer to 1, the more information is retained), is the forget gate weight matrix, is the hidden layer output at the previous moment With the current input The concatenation vector of Input features for the current moment, is the bias of the forget gate, ( ) is the Sigmoid activation function; The input gate and candidate cell state are used to determine the input and update of new information at the current moment , ,in, is the output of the input gate at time t, which is used to control the proportion of new information input; is the candidate cell state at time t, 、 are the weight matrices corresponding to the input gate and cell state, 、 are the biases corresponding to the input gate and cell state, is the hyperbolic tangent activation function; Cell state update, used to integrate historical information with current new information ,in represents the cell state at time t, is the cell state at the previous moment, is element-wise multiplication; Output gate: determines the output of the current cell state , ,in, is the output of the output gate at time t, is the hidden layer output vector at the current moment, is the weight matrix of the output gate, is the bias of the output gate; The remaining service life calculation subunit extracts long-term dependency features through the fully connected layer and fits the component degradation amount. The time series evolution curve (specifically, the degradation amount is mapped by the fully connected layer , is calculated for the fully connected layer), combined with the preset failure threshold , the predicted remaining service life is calculated according to the following formula : ,in is the degradation rate, The preset credit life. is the preset collection interval; Fault risk assessment subunit, calculates health based on degradation , the formula is Set five levels of fault warning, including level 1: no warning, level 2: slight warning, level 3: moderate warning, level 4: severe warning, and level 5: emergency warning. Match the health level with the five warning levels to output the predicted fault warning level F of the corresponding component. The model optimization subunit uses the Adam optimizer and the joint loss function: , perform parameter optimization, where MSE is the mean square error and CE is the cross entropy loss.

[0028] It should be noted that by refining the functions and algorithm logic of each sub-unit of the predictive maintenance unit, standardizing input features to ensure data consistency, improving the LSTM network to accurately capture timing features, combining the fully connected layer and formula calculation to achieve remaining life prediction and five-level fault warning, and the Adam optimizer and joint loss function to improve model accuracy, the overall accurate prediction of component status and continuous optimization of the model are achieved, providing reliable support for the system's proactive maintenance and enhancing operational safety and stability.

[0029] In the present invention, the dynamic adjustment model includes a road condition recognition sub-model and a power optimization sub-model, specifically: The road condition recognition sub-model identifies the current road type, slope, and congestion status based on the received road condition information. Specifically: Identify road types and obtain road condition information, including vehicle speed, acceleration, braking frequency, and vehicle distance; extract feature vectors from road condition information , including the average speed within the set time window , speed variance Average distance to the vehicle ahead and braking frequency ; Calculate the probability of each road type through the softmax classifier , the formula is: ,in is the classification weight matrix (corresponding to three types of roads, including urban roads / highways / mountain roads), is the classification bias vector; the category corresponding to the maximum probability of each road type is taken as the road condition recognition result; Slope calculation, based on the vehicle dynamics balance equation, combines acceleration and resistance parameters to calculate the road slope , the formula is ,in is the vehicle acceleration, m is the vehicle mass, Expressed as vehicle driving resistance (the calculation formula is , represents the air density, represents the air resistance coefficient, A represents the frontal area of ​​the vehicle), g is the acceleration due to gravity; Congestion status determination: quantify the current traffic status through the congestion index and calculate the traffic status evaluation value , the formula is ;in is the free flow speed of the current road; Preset congestion status groups, including smooth, slow, and congested; compare the traffic status evaluation value with the preset congestion status group to output the corresponding congestion status; The power optimization sub-model dynamically adjusts the output voltage and current of the DCDC converter based on the road condition recognition results and battery status information. Specifically: The benchmark power is calculated based on the current low-voltage load demand and the basic battery status. The formula is: ;in is the reference output power, is the actual power demand of the current low-voltage electrical equipment, is the standby power factor, SOC is the battery state of charge; Calculate the road condition correction coefficient and dynamically correct the benchmark power according to the road slope and congestion status. The expression is: ;in 、 are the uphill and downhill correction coefficients, is the congestion correction factor; It should be noted that the standby power factor , Uphill and downhill correction coefficient 、 Correction factor for battery status The specific values ​​need to be determined by those skilled in the art through routine experiments; Battery state correction coefficient calculation, combined with battery health status and temperature parameters for secondary correction, the expression is ; Where SOH is the battery health status, is the current temperature of the battery, The optimal operating temperature for the battery is 、 、 are the correction coefficients of the corresponding parameters respectively; The target output parameters are determined, and the final output parameters are calculated based on the above correction results. The final output parameters include the target output power , target output voltage , target output current ,expression: , , ; in represents the reference output voltage, Indicates the correction factor of battery health status to output voltage; During the adjustment process, the change rate of each output parameter is calculated and the maximum allowable change threshold is preset. If the change rate is greater than the preset maximum allowable change threshold, the output parameter corresponding to the preset allowable change limit value is taken to ensure the smoothness and stability of power adjustment.

[0030] It should be noted that the road condition identification sub-model is used to accurately judge the road type, slope and congestion status, and the output parameters are dynamically adjusted in combination with the multi-level correction mechanism (baseline power + road condition correction + battery status correction) of the power optimization sub-model. At the same time, the smoothness of the adjustment is guaranteed by the preset change threshold, which not only achieves the precise adaptation of energy distribution to driving conditions and battery status, but also improves the stability of power regulation and enhances the energy efficiency and adaptability of the system.

[0031] In the present invention, the control execution module includes a power regulation unit, a protection control unit and a mode switching unit, which are specifically as follows: The power regulation unit is used to adjust the output power of the DCDC converter, specifically: It consists of a digital signal processor and a drive circuit, and achieves dynamic output power matching by closed-loop adjustment of the power device switching duty cycle. The process is as follows: Collect the measured voltage and output current at the DCDC output end and calculate the actual output power , the formula is: ; Combined with the target power output of the machine learning module , calculate the power deviation: ; Then use the PI control algorithm to generate the duty cycle correction value, the formula is: ; Among them is Proportional coefficient, is the integration coefficient; The current duty cycle is recorded as , and is updated by the duty cycle correction value according to the following formula: ; Define constraint rules, including boundary constraints and step constraints; among which boundary constraints , Indicates the maximum threshold of duty cycle; step size constraint: ; The protection control unit performs overvoltage, overcurrent, and overheat protection actions based on the output of the predictive maintenance unit and real-time electrical parameters and physical state parameters. The specific process is as follows: Set the overvoltage threshold, overcurrent threshold, and overheat threshold respectively; the overvoltage threshold is calculated by the formula , Indicates the rated voltage corresponding to the component, Indicates the overvoltage coefficient, which is used to reserve a certain safety margin on the voltage; the overcurrent threshold is calculated by the formula , Indicates the rated current corresponding to the component, Indicates the overcurrent coefficient, which is used to reserve a certain safety margin on the current; the overheating threshold is calculated by the formula , Indicates the rated operating temperature of the component. Indicates superheat temperature difference; The protection is triggered when any of the following conditions is detected and meets the preset time continuously: , then the corresponding component indicates output overvoltage; , then the corresponding component indicates output overcurrent; or power device temperature , indicating that the device is overheated and the duration is greater than the preset continuous threshold; Combined with the fault warning level F output by the predictive maintenance unit, hierarchical protection actions are performed: When the fault warning level is level one, no operation is performed; When the fault warning level is level 2, the output power is reduced to the preset level 1 ratio (e.g. 80%) of the rated value, and the supporting cooling system is operated at the preset low speed; When the fault warning level is level three, the output power is reduced to the rated value of the preset level two ratio (such as 50%), and the supporting cooling system is operated at the preset high speed; When the fault warning level reaches level four, the preset non-critical load output will be shut down, and only the minimum voltage of the power system will be maintained; When the fault warning level reaches level five, the main circuit relay is immediately cut off, triggering a vehicle fault alarm, which is displayed on the vehicle's dashboard and the data is uploaded to the cloud; The mode switching unit is used to realize automatic switching of charging and discharging modes and power levels, specifically: Define the mode switching judgment conditions: Charging mode: When , Indicates the low battery threshold. When an external AC power source is detected, the system switches to charging mode and executes the constant current / constant voltage charging strategy. Standby mode: When , Indicates the battery is fully charged and when there is no low-voltage load power demand, it switches to standby mode; Energy recovery mode: When a braking signal is detected (the vehicle controller outputs the braking level) and When the brake is turned on, it switches to energy recovery mode, converting the braking kinetic energy into electrical energy and feeding it back to the battery; Indicates the maximum power threshold allowed for battery recycling; Mode switching transition control: Specifically, a linear power transition algorithm is used to avoid voltage / current shock during mode switching. The formula is: td; where Indicates the output power before switching, Indicates the output power of the target mode, td indicates the current transition time, Indicates the mode switching transition time, and , used to ensure stable output during the switching process without obvious voltage fluctuations; Mode status is maintained, and the corresponding status register is set for each mode to record the upper and lower limits of output voltage, upper and lower limits of output current and protection threshold adaptation.

[0032] It should be noted that by clarifying the specific implementation logic of each unit of the control execution module, the power regulation unit dynamically matches the output power through closed-loop control, the protection control unit performs hierarchical protection in combination with the fault warning level, and the mode switching unit achieves smooth switching through judgment conditions and linear transition algorithms. The three work together to ensure the accuracy of power regulation, the gradient of system protection and the stability of mode switching, and comprehensively improve the reliability and adaptability of equipment operation.

[0033] In the present invention, a storage module is further included for storing historical electrical parameters and physical states, model parameters and fault records; It should be noted that the historical electrical parameters and physical states include the voltage feature subset (input voltage, high and low voltage output voltage, etc.), current feature subset (input and output current, branch current, etc.), temperature feature subset (power devices, inductance, ambient temperature, etc.), vibration feature subset (three-axis acceleration and effective value), and insulation resistance feature subset collected by the sensor module; the model parameters cover the weight matrix, bias, correction coefficient, etc. of the prediction model and dynamic adjustment model in the machine learning module, as well as the parameters after iterative optimization of the model update unit; the fault record includes the fault warning level output by the predictive maintenance unit, the overvoltage / overcurrent / overheating and other fault types triggered by the protection control unit, the occurrence time, and treatment measures.

[0034] The above formulas are calculated using dimensionless processing (such as standardization), and only numerical values ​​are used in the calculations. The formulas are based on a large amount of measured data and optimized through software simulation to fit actual working conditions. The preset parameters in the formulas are set by technical personnel in this field according to actual needs.

[0035] This solution can be implemented through software, hardware, or a combination thereof. The relevant program can be stored on a computer-readable medium (such as a USB flash drive, hard disk, optical disk, etc.), and when loaded and executed, the aforementioned functions can be implemented.

[0036] It should be noted that the execution order of each implementation step is determined by the functional logic and does not constitute a limitation; the units in this article are divided into logical functions and can be adjusted according to actual needs.

[0037] Other embodiments of the present invention will readily occur to those skilled in the art after considering the specification and practicing the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein. The description and examples are to be considered as exemplary only, with the true scope and spirit of the invention being indicated by the following claims.

[0038] It should be understood that the present invention is not limited to the exact construction described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

Claims

1. New energy vehicle intelligent DCDC converter integrated charger, characterized by: include: The main circuit module integrates the on-board charger unit, DCDC conversion unit and high-voltage power distribution unit; The sensor module includes a plurality of sensor groups, which are respectively arranged on each key component of the main circuit module and are used to collect electrical parameters and physical state parameters of the corresponding components; a data processing module connected to the sensor module, configured to receive and pre-process the electrical parameters and physical state parameters; Preprocessing includes filtering, calibration, normalization and feature extraction; A machine learning module, connected to the data processing module, having a built-in trained prediction model and dynamic adjustment model, for analyzing and calculating preprocessed electrical parameters and physical state parameters; a control execution module, connected to the machine learning module and the main circuit module respectively, and configured to control the operating state of the main circuit module according to the output of the machine learning module; The communication module is used to realize data interaction with the vehicle controller and battery management system, and receive road condition information and battery status information.

2. The new energy vehicle intelligent DCDC converter integrated charger according to claim 1, characterized in that: The sensor module includes: Voltage sensor group, used to collect input voltage, output voltage and voltage of each key node; Current sensor group, used to collect input current, output current and current of each branch; Temperature sensor group, used to collect the temperature of power devices, inductor temperature and ambient temperature; Vibration sensors, used to collect vibration parameters of key components; Insulation monitoring sensor, used to monitor the insulation status of the system.

3. The new energy vehicle intelligent DCDC converter integrated charger according to claim 1, characterized in that: The machine learning module includes: A predictive maintenance unit, which predicts the remaining service life and potential failure risk of each component through the prediction model based on the preprocessed electrical parameters and physical state parameters; A dynamic energy allocation unit, combining the road condition information and the battery status information, generates a DCDC output power adjustment strategy through the dynamic adjustment model; The model updating unit is used to perform online optimization on the prediction model and the dynamic adjustment model according to actual operation data.

4. The new energy vehicle intelligent DCDC converter integrated charger according to claim 3, characterized in that: The predictive maintenance unit includes: The input feature processing subunit is used to standardize the preprocessed time series parameters; LSTM prediction network, used to process time series features through an improved gating mechanism to achieve time series analysis of component status; The remaining service life calculation subunit is used to calculate and predict the remaining service life based on the component degradation evolution curve and failure threshold; The fault risk assessment subunit is used to calculate the health level based on the degradation amount and output the fault warning level according to the five-level fault warning level; The model optimization subunit is used to optimize the prediction model parameters using the Adam optimizer and the joint loss function.

5. The new energy vehicle intelligent DCDC converter integrated charger according to claim 3, characterized in that: The dynamic adjustment model includes a road condition recognition sub-model and a power optimization sub-model: The road condition identification sub-model identifies the current road type, slope and congestion status based on the received road condition information; The power optimization sub-model dynamically adjusts the output voltage and current of the DCDC converter according to the road condition recognition result and the battery status information.

6. The new energy vehicle intelligent DCDC converter integrated charger according to claim 4, characterized in that: The control execution module includes: A power regulation unit, used to regulate the output power of the DCDC converter; a protection control unit, which performs overvoltage, overcurrent, and overheat protection actions based on the output of the predictive maintenance unit and real-time electrical parameters and physical state parameters; The mode switching unit is used to realize automatic switching of charging and discharging modes and power levels.

7. The new energy vehicle intelligent DCDC converter integrated charger according to claim 1, characterized in that: The system also includes a storage module for storing historical electrical parameters and physical states, model parameters and fault records.

8. The new energy vehicle intelligent DCDC converter integrated charger according to claim 6, characterized in that: The protection control unit performs hierarchical protection actions based on the fault warning level: When the fault warning level is level 2, the output power is reduced to the rated value of the preset level 1 ratio, and the cooling system is controlled to run at the preset low speed; When the fault warning level is level three, the output power is reduced to the rated value of the preset level two ratio, and the cooling system is controlled to run at the preset high speed; When the fault warning level is level 4, the preset non-critical load output is turned off; When the fault warning level reaches level five, the main circuit relay is cut off, triggering the vehicle fault alarm and uploading data to the cloud.

Images