Fuel cell system domain controller FDCU and control method

Abstract

The invention discloses a fuel cell system domain controller FDCU and a control method, and relates to the field of fuel cell system control, and the control method comprises the following steps: S1, carrying out the multi-dimensional grouping of power supply channels, and decomposing a complex multi-port power flow into a simplified multi-port / two-port converter topology; s2, the running state of the B0P component is monitored in real time, and the change trend of the power requirement of the component is pre-judged; s3, selecting an optimal grouping scheme according to the change of the power demand of the B0P component, adjusting the grouping of the power supply channels in real time, and adjusting the topological connection mode of the converter; and S4, each power supply channel is controlled through PID, the output voltage / current is adjusted in real time, and the grouping strategy weight is dynamically adjusted according to the emergency degree of the system. According to the invention, comprehensive improvement of power supply efficiency, control stability, flexibility and cost of a fuel cell system is realized through converter topology optimization technical means of multi-dimensional grouping, dynamic adjustment of a power supply strategy and reconstruction of a power flow path.

Description

Technical Field

[0001] The present invention relates to the field of fuel cell system control technology, and in particular to control or regulation, and specifically to a fuel cell system domain controller FDCU and a control method. Background Art

[0002] As an efficient and environmentally friendly energy conversion device, the fuel cell system has broad application prospects in the fields of transportation, distributed power generation, etc. A typical fuel cell system is mainly composed of a fuel cell stack, BOP components, and a vehicle high-voltage system. Among them, BOP components, including but not limited to air compressors, hydrogen circulation pumps, cooling systems, humidifiers, etc., are key auxiliary equipment to ensure the normal operation of the fuel cell system.

[0003] Due to the wide variety of BOP components and their widely varying working characteristics, how to effectively manage and coordinate the power requirements of these components has become an urgent problem to be solved. Traditionally, the method of designing a separate power supply channel for each BOP component not only increases the complexity and cost of the system, but also makes it difficult to achieve flexible and efficient power distribution and management. Especially under dynamic conditions, such as vehicle acceleration and deceleration, the power requirements of BOP components will change rapidly, which puts higher requirements on the flexibility and response speed of power supply.

[0004] Therefore, it is necessary to improve the deficiencies in the prior art to solve the above problems. Summary of the invention

[0005] The present invention overcomes the deficiencies of the prior art and provides a fuel cell system domain controller FDCU and a control method.

[0006] To achieve the above object, the technical solution adopted by the present invention is: a control method of a fuel cell system domain controller FDCU, comprising the following steps: S1. Based on the power requirements and electrical characteristics of BOP components, multi-dimensional grouping of power supply channels is performed to determine the demand boundary of each group. According to the requirements and characteristics of each group, the complex multi-port power flow is decomposed into simplified multi-port / two-port converter topologies; S2. Real-time monitoring of the operating status of BOP components, combined with historical data and operating condition models, to predict the changing trend of component power demand; S3. According to the change of power demand of BOP components, the optimal grouping scheme is selected through the transformation strategy, the grouping of power supply channels is adjusted in real time, and the converter topology connection mode is adjusted to match the power flow demand of the new grouping; S4. Control each power supply channel through PID, adjust the output voltage / current in real time, compensate for load fluctuations, and dynamically adjust the grouping strategy weight according to the system urgency to optimize the selection of the optimal power supply strategy.

[0007] In a preferred embodiment of the present invention, the step S1 includes the following sub-steps: S11, establishing a multi-dimensional grouping parameter matrix, including the power level, operating voltage range, current fluctuation characteristics, response time requirements and operating frequency parameters involved; S12, based on a clustering algorithm, merging BOP components with similar parameters into the same power supply group; Objective function: Minimize the within-group squared error: ; in, is the number of groups; It is Group; It is Group center point; Grouping rules: Similarity is measured by Euclidean distance: ; in, and They are the feature vectors of two different BOP components in parameter space; and The components are and components In the Specific values ​​of the parameters; S13. Establish a dynamic demand boundary model for each power supply group, including the maximum power threshold, voltage tolerance range, and current fluctuation limit involved; Maximum power threshold , based on the maximum instantaneous power in the group and the safety margin: ; in, is the power safety factor; is the maximum instantaneous power in the group, ,in, It is a component Rated power, is the instantaneous load factor; Voltage tolerance range , based on the dynamic expansion of rated voltage and fluctuation range: ; in, Yes Group Rated voltage; is the voltage tolerance factor; Current fluctuation limitation , set the threshold based on the standard deviation of current fluctuation: ; in, Yes Group The average current of is the standard deviation of the current fluctuation within the group, , is the number of components within the group; S14. Based on the power flow path reconstruction method, a simplified multi-port / two-port converter topology is designed to optimize the power flow path. The input-output relationship of the multi-port converter is expressed as: ; in, is the number of input ports; is the number of output ports; It is The power of each input port; It is The power of each output port; It is from Input port to The transmission efficiency of each output port; The Scattering parameter matrix of the two-port converter is expressed as: ; in, and are the reflection coefficients of the input and output ports, respectively; and is the transmission coefficient, which is used to describe the power transfer characteristics between ports.

[0008] In a preferred embodiment of the present invention, in the step S11, the relevant parameters of all BOP components are collected, and for each BOP component, the parameters are defined and quantified: the power level is divided into levels according to the rated power; the operating voltage range is based on the ratio of the voltage range width to the rated voltage; the current fluctuation characteristics are based on the ratio of the current fluctuation standard deviation to the average current; the response time requirement is the normalized response time; the operating frequency parameter is based on the offset ratio between the operating frequency and the base frequency; For the multidimensional grouping parameter matrix, the weight vector is multiplied by the parameter matrix to obtain a weighted matrix : ; in, is a multidimensional grouping parameter matrix; is the transpose of the weight vector; is the power level; It is a component The weight vector of .

[0009] In a preferred embodiment of the present invention, the step S2 includes the following sub-steps: S21, real-time monitoring of dynamic power requirements, electrical parameter deviations and operation mode switching signals of BOP components; S22. Predict the power demand change trend of BOP components using the operating condition model; Operating model combining physical equations and data-driven models: ; in, is the weight coefficient; It is a predicted value based on the physical model; is the machine learning output; S23, predicting the triggering threshold of grouping adjustment through the grouping decision model trained with historical data; The group decision model is quickly deployed through logistic regression: ; in, It's at the time , the probability of triggering group adjustment, range ; is the intercept term of the model; It is Features The regression coefficient of .

[0010] In a preferred embodiment of the present invention, the step S3 includes the following sub-steps: S31, when the power or electrical parameters of the monitored BOP components exceed the current grouping threshold, dynamic grouping reorganization is triggered; Model output trigger probability, set dynamic threshold based on historical data distribution : ; in, is the recent average trigger probability; is the standard deviation; is the sensitivity coefficient; S32, selecting the optimal grouping scheme through a decision tree algorithm and reallocating power supply channels; The decision tree model selects split features and thresholds to minimize the Gini impurity of child nodes: ; in, is the total number of categories in the dataset; It is the first A subset of samples of each category; is the current data set to be segmented; S33, switch the parallel / series mode of the converter topology, use soft switching technology to reduce transient impact, and calibrate the topology parameters of the reorganized power supply channel to ensure that the power flow matches the demand.

[0011] In a preferred embodiment of the present invention, in the step S33, for topology switching, when the original parallel power supply channel is switched to the series mode: ; in, is the output voltage of the new topology after switching; and is the output voltage of the original parallel power supply channel; is the output current of the new topology after switching; and is the output current of the original parallel power supply channel; Add zero voltage switching condition during switching: ; in, is the resonant inductance value, used to achieve ZVS; is the resonant tank peak current; is the output capacitance of the switch tube; is the drain-source voltage of the switch tube; Adjust the converter inductor / capacitor values ​​so that the output matches the power flow requirements of the new group: ; in, is the calibrated inductance value; is the converter input voltage; is the duty cycle; is the allowable current ripple; is the switching frequency; Check whether the reorganized power flow satisfies: ; in, is the actual measured power; is the target power; is the allowable relative error.

[0012] In a preferred embodiment of the present invention, the step S4 includes the following sub-steps: S41. Design an independent PID controller for each power supply channel to adjust the output voltage / current in real time to compensate for load fluctuations; Continuous domain PID control law: ; in, is the control output; is the error signal, or ,in, is the voltage setting value, is the actual output voltage, is the current setting value, is the actual output current; is the cumulative amount of error; is the rate of change of the error; , and are the proportional, integral and differential coefficients respectively; is the current time, indicating the real-time running time of the controller; is the integral variable, which represents the time integral process from the initial moment to the current moment; S42, setting strategy priorities according to the temperature and voltage safety thresholds of the battery stack, and dynamically adjusting the grouping weights; S43, selecting a power supply strategy by optimizing a cost function, wherein the objective function includes minimum energy consumption or minimum temperature rise; The objective function to minimize energy consumption is: ; in, For Channel The total loss, It is a channel The effective value of the current, It is a channel The equivalent resistance of is the switching frequency, is the switching loss energy in a single switching cycle; is the penalty term for power deviation from the rated value, It is a channel The actual output power, It is a channel Rated power; is the power deviation weight coefficient; is the total number of power supply channels in the system; The objective function to minimize the temperature rise is: ; in, For Channel The temperature rise, It is a channel The thermal resistance, It is a channel The effective cooling area; is the penalty term for the temperature rise rate; is the weight coefficient of temperature rise rate; Multi-objective cost function: ; in, is the weight of energy consumption and temperature rise.

[0013] In a preferred embodiment of the present invention, in the step S42, the stack temperature is defined as and the voltage is safe Threshold: Safety range of stack temperature threshold: , for emergency scoring: ; Safety range of voltage safety threshold: , for emergency scoring: ; Calculate the comprehensive emergency priority by weighted comprehensive score: ; in, and are the weights of temperature and voltage respectively; Priority mapping: Low urgency: maintain the default group weight; Medium urgency: adjust the weight of the key group proportionally; High urgency: forcibly increase the weight of the key group to the upper limit; Dynamically adjust group weights: ; in, It is a group The base weight of It is a group emergency response factor.

[0014] The present invention provides a fuel cell system domain controller FDCU, which is used to implement any of the control methods described above, including: A multi-dimensional grouping module, used to perform multi-dimensional grouping according to the power requirements and electrical characteristics of BOP components and generate an adaptive simplified converter topology; Condition monitoring module, used to monitor the operating parameters of BOP components in real time; Prediction module, used to predict the power demand change trend in combination with the operating condition model; A dynamic adjustment module is used to trigger group reorganization according to the prediction results and adjust the converter topology connection mode; PID control module, used to independently adjust the output of each power supply channel and dynamically optimize the grouping strategy weight.

[0015] The present invention provides a fuel cell system, comprising the FDCU described above, and a BOP component, a fuel cell stack and a whole vehicle high-voltage system connected to the FDCU.

[0016] The present invention solves the defects existing in the background technology and has the following beneficial effects: The present invention provides a fuel cell system domain controller FDCU and a control method. Through a multi-dimensional grouping strategy, BOP components with similar power requirements and electrical characteristics are classified into the same power supply group, which can optimize the power flow path, reduce conversion losses, and improve power supply efficiency. The real-time monitoring and prediction mechanism is combined with independent PID control and dynamic weight adjustment to enhance the system's ability to compensate for load fluctuations and stability under different working conditions, thereby achieving efficient, stable, flexible and economical operation of the fuel cell system power supply efficiency.

[0017] In the present invention, through a multi-dimensional grouping strategy, BOP components with similar power requirements and electrical characteristics are classified into the same power supply group, and combined with a method of reconstructing the power flow path, the complex multi-port power flow is simplified into an optimized two-port / multi-port converter topology, which effectively reduces the energy loss in the power conversion process, shortens the power transmission path, and significantly improves the overall power supply efficiency.

[0018] In the present invention, by real-time monitoring of the dynamic power demand, electrical parameter offset and operation mode switching signal of the BOP components, combined with the operating condition model to predict the power demand change trend, the power supply channel grouping and converter topology can be adjusted in advance to avoid voltage fluctuations caused by instantaneous power shocks. At the same time, an independent PID controller is designed for each power supply channel, which can dynamically adjust the grouping strategy weight according to the system urgency, giving priority to ensuring the power supply stability of key components. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art are briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work. Figure 1 It is a flow chart of a control method of a fuel cell system domain controller FDCU according to a preferred embodiment of the present invention; Figure 2 It is an electrical schematic diagram of a fuel cell system domain controller FDCU of a preferred embodiment of the present invention; Figure 3 It is a schematic diagram of the structure of a fuel cell system domain controller FDCU according to a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0022] like Figure 1 and Figure 2 As shown, a control method of a fuel cell system domain controller FDCU comprises the following steps: S1. Based on the power requirements and electrical characteristics of BOP components, multi-dimensional grouping of power supply channels is performed to determine the demand boundary of each group. According to the requirements and characteristics of each group, the complex multi-port power flow is decomposed into simplified multi-port / two-port converter topologies; S2. Real-time monitoring of the operating status of BOP components, combining historical data with operating condition models, and predicting the changing trends of component power requirements and electrical characteristics; S3. According to the changes in the power requirements and electrical characteristics of the BOP components, the optimal grouping scheme is selected through the transformation strategy, the grouping of the power supply channels is adjusted in real time, and the converter topology connection mode is adjusted to match the power flow requirements of the new grouping; S4. Control each power supply channel through PID, adjust the output voltage / current in real time, compensate for load fluctuations, and dynamically adjust the grouping strategy weight according to the system urgency to optimize the selection of the optimal power supply strategy.

[0023] It should be noted that the present invention achieves a comprehensive improvement in the power supply efficiency, control stability, flexibility and cost of the fuel cell system through multi-dimensional grouping, dynamic adjustment of power supply strategy, and reconstruction of power flow path converter topology optimization technology. It not only improves the overall performance of the system, but also provides more solid technical support for the widespread application of fuel cell systems, and promotes the further development and commercialization of fuel cell technology.

[0024] In some specific embodiments, the step S1 includes the following sub-steps: S11, establishing a multi-dimensional grouping parameter matrix, including the power level, operating voltage range, current fluctuation characteristics, response time requirements and operating frequency parameters involved; S12, based on a clustering algorithm, merging BOP components with similar parameters into the same power supply group; S13. Establish a dynamic demand boundary model for each power supply group, including the maximum power threshold, voltage tolerance range, and current fluctuation limit involved; S14. Based on the power flow path reconstruction method, a simplified multi-port / two-port converter topology is designed to optimize the power flow path.

[0025] In this embodiment, in step S11, relevant parameters of all BOP components are collected, and parameters are defined and quantified for each BOP component: Power level , divided into grades according to rated power: ; Operating voltage range , based on the voltage range width (difference between maximum and minimum) and the rated voltage Ratio of: ; For example: the rated voltage of a component is 24V, and the operating voltage range is 20-28V. ; Current fluctuation characteristics , based on the current fluctuation standard deviation With the average current Ratio of: ; in, ; is the number of sampling points; is the instantaneous current value; Response time requirements ,Normalized response time (time required from 0 to 90% of the steady-state value): ; in, The maximum response time allowed by the system; for example: The maximum response time allowed by the system , the response time of a component ,but ; Working frequency parameters , based on the operating frequency With base frequency The offset ratio is: ; For example: the operating frequency of a component , fundamental frequency ,but ; In order to eliminate the influence of different parameter dimensions, the parameters of different dimensions are standardized to the interval [0,1] to facilitate matrix operations: ; in, is the original parameter value; and are the minimum and maximum values ​​of this parameter in all BOP components, respectively; Arrange the normalized parameters of all BOP components in rows to form a matrix : ; Among them, the matrix dimensions are: ; is the total number of BOP parts; Assign weight vectors to different parameters according to system requirements : ; Among them, the sum of weights is 1, and the weight values ​​are determined by expert evaluation or analytic hierarchy process (AHP); Multiply the weight vector by the parameter matrix to get the weight matrix : ; in, is a multidimensional grouping parameter matrix; is the transpose of the weight vector, with the dimension changed from 1×5 to 5×1 to facilitate matrix multiplication with the parameter matrix; is the power level (for the operating voltage range, current fluctuation characteristics, response time requirements and operating frequency parameters, just change the parameters accordingly); It is a component The weight vector of Finally, a comprehensive score of each BOP component is obtained for clustering.

[0026] In this embodiment, in step S12, a clustering algorithm is used to group the weighted matrix: Objective function: Minimize the within-group squared error: ; in, is the number of groups; It is Group; It is Group center point; Grouping rules: Similarity is measured by Euclidean distance: ; in, and They represent the eigenvectors of two different BOP components (such as air compressor, hydrogen pump, cooling water pump, etc.) in the parameter space. Each eigenvector contains the five key parameters of the component (from the multidimensional grouping parameter matrix of step S11): :Power level, : Operating voltage range, :Current fluctuation characteristics, : Response time requirements, : Operating frequency parameters; and The components are and components In the The specific value of a parameter.

[0027] In this embodiment, in step S13, the maximum power threshold , based on the maximum instantaneous power in the group and the safety margin: ; in, is the power safety factor, which is 0.1~0.2; is the maximum instantaneous power in the group, ,in, It is a component Rated power, is the instantaneous load factor (0~1); Dynamic adjustment: real-time update based on load rate changes: ; in, is the change in the average load factor within the group; is the base load factor; is the adjustment factor (0.05); Voltage tolerance range , based on the dynamic expansion of rated voltage and fluctuation range: ; in, Yes Group Rated voltage; is the voltage tolerance factor (0.1); Dynamic adjustment: Correction based on real-time voltage fluctuation rate: ; in, is the within-group voltage standard deviation; is the average voltage within the group; Current fluctuation limitation , set the threshold based on the standard deviation of current fluctuation: ; in, Yes Group The average current of is the standard deviation of the current fluctuation within the group, , is the number of components within the group; Dynamic adjustment: Optimize according to load mutation frequency: ; in, is the number of load mutations per unit time; is the adjustment factor (0.02).

[0028] In this embodiment, in step S14, a mathematical model of power flow is established to describe the transmission relationship and loss of power between different components: ; in, is the output power; is the input power; is the efficiency of the converter, and its value is affected by circuit component parameters and operating conditions; For modular design: divide the converter into multiple functional modules, such as boost module, buck module, isolation module, etc. Each module has clear power processing functions and interface specifications; For path optimization: determine the optimal power flow path based on the needs and characteristics of the power supply group; for example, for a power supply group with high power demand and high voltage requirements, give priority to the combined path of the boost module and the isolation module; for a power supply group with low power and stable voltage, a simple buck module path is sufficient; For component sharing and reuse: identify sharable components, such as magnetic components and capacitors, and reuse them in multiple power flow paths to reduce the number and size of components and improve system integration; In order to meet the needs of multiple power supply groups, a multi-port converter topology is designed so that each port corresponds to one or more power supply groups. The input-output relationship of the multi-port converter is expressed as: ; in, is the number of input ports; is the number of output ports; It is The power of each input port; It is The power of each output port; It is from Input port to The transmission efficiency of each output port; The Scattering parameter matrix of the two-port converter is expressed as: ; in, and are the reflection coefficients of the input and output ports, respectively; and is the transmission coefficient, which is used to describe the power transfer characteristics between ports.

[0029] In some specific embodiments, the step S2 includes the following sub-steps: S21, real-time monitoring of dynamic power requirements, electrical parameter deviations and operation mode switching signals of BOP components; S22. Predict the power demand change trend of BOP components using the operating condition model; S23. Predict the trigger threshold of group adjustment through the group decision model trained by historical data.

[0030] In this embodiment, in step S21, the dynamic power demand of the BOP components (such as power fluctuation when the air compressor speed suddenly changes), electrical parameter offset (voltage / current offset, frequency response) and operation mode switching signal operation mode switching (startup, steady state, shutdown) are monitored in real time through sensors.

[0031] In this embodiment, in step S22, the system operating state is divided into typical operating conditions: cold start, steady-state operation, acceleration, and shutdown; a unique identifier is assigned to each operating condition: cold start operating condition , Steady-state motion condition , transient load conditions , and establish a correlation table between the operating mode and key parameters, such as shown in Table 1; Table 1:

[0032] Operating model combining physical equations and data-driven models: ; in, is the weight coefficient; It is a predicted value based on the physical model; is the machine learning output; For real-time prediction and dynamic correction: collect the current working parameters (such as temperature , Stack voltage ), and match it to the corresponding working mode ; The exponential growth model is used to describe the power climbing process: ; in, is the steady-state power; is the time constant (fitted by historical data); Adjust the forecast value according to the real-time error: ; in, is the BOP component power requirement; is the proportional gain.

[0033] In this embodiment, in step S23, the group decision model is quickly deployed through logistic regression: ; in, It's at the time , the probability of triggering group adjustment, range ; is the intercept term of the model, which means that when all features The base logarithmic probability of the trigger probability when ; It is Features The regression coefficient reflects the direction and intensity of the impact of this feature on the trigger probability.

[0034] In some specific implementations, the step S3 includes the following sub-steps: S31, when the power or electrical parameters of the monitored BOP components exceed the current grouping threshold, dynamic grouping reorganization is triggered; S32, selecting the optimal grouping scheme through a decision tree algorithm and reallocating power supply channels; S33, switch the parallel / series mode of the converter topology, use soft switching technology to reduce transient impact, and calibrate the topology parameters of the reorganized power supply channel to ensure that the power flow matches the demand.

[0035] In this embodiment, in step S31, the model outputs the trigger probability , set dynamic thresholds based on historical data distribution : ; in, is the recent average trigger probability; is the standard deviation; is the sensitivity factor (2.0); Calculate the true positive rate (TPR) and false positive rate (FPR) at different thresholds, and select the threshold that maximizes the Youden index: ; According to the current working mode Match the predefined threshold table, see Table 2; Table 2:

[0036] In this embodiment, in step S32, when selecting the optimal grouping, the real-time parameters of the current BOP components are used as input data, and a set of candidate grouping schemes is generated based on parameter similarity and hardware topology constraints. ; Each candidate grouping scheme Features include: Power matching: total power demand within the group and power supply channel capacity Ratio of: ; in, is power; Voltage fluctuation rate: the maximum voltage deviation within the group and the rated voltage Ratio of: ; in, is the voltage; Current balance: standard deviation of current within the group With the average current Ratio of: ; Calculate the overall score for each candidate solution : ; in, , and is the weight coefficient, satisfying ; The feature vector of the candidate solution Input the decision tree model to predict the feasibility label. The decision tree model selects split features and thresholds to minimize the Gini impurity of the child nodes: ; in, is the total number of categories in the dataset; It is the first A subset of samples of each category; is the data set to be segmented. For example, in the BOP component grouping problem, It can represent the set of all components to be grouped; Select the grouping scheme with the highest score and “feasible” feasibility to complete the redistribution of power supply channels.

[0037] In this embodiment, in step S33, for topology switching, when the original parallel power supply channel is switched to the series mode: ; in, is the output voltage of the new topology after switching; and is the output voltage of the original parallel power supply channel; is the output current of the new topology after switching; and is the output current of the original parallel power supply channel; Adding zero voltage switching (ZVS) conditions to the switching process: ; in, is the resonant inductance value, used to achieve ZVS; is the resonant tank peak current; is the output capacitance of the switch tube; is the drain-source voltage of the switch tube; Adjust the converter inductor / capacitor values ​​so that the output matches the power flow requirements of the new group: ; in, is the calibrated inductance value; is the converter input voltage; is the duty cycle ( ); is the allowable current ripple; is the switching frequency; Check whether the reorganized power flow satisfies: ; in, is the actual measured power; is the target power; is the allowable relative error.

[0038] In some specific embodiments, the step S4 includes the following sub-steps: S41. Design an independent PID controller for each power supply channel to adjust the output voltage / current in real time to compensate for load fluctuations; S42, setting strategy priorities according to the temperature and voltage safety thresholds of the battery stack, and dynamically adjusting the grouping weights; S43. Select a power supply strategy by optimizing the cost function, where the objective function includes minimum energy consumption or minimum temperature rise.

[0039] In this embodiment, in step S41, the control variable selection of the PID controller is: voltage control mode (with output voltage as the control target, suitable for scenarios with high voltage regulation requirements) and current control mode (with output voltage For control purposes, it is suitable for dynamic load scenarios); Continuous domain PID control law: ; in, is the control output; is the error signal, (voltage control mode) or (current control mode), where, is the voltage setting value, is the actual output voltage, is the current setting value, is the actual output current; is the cumulative amount of error; is the rate of change of the error; , and are the proportional, integral and differential coefficients respectively; is the current time, indicating the real-time running time of the controller; is the integral variable, which represents the time from the initial moment ( ) to the current moment ( ) is a time integration process.

[0040] In this embodiment, in step S42, the stack temperature is defined as and voltage safety Threshold: Safety range of stack temperature threshold: , for emergency scoring: ; Safety range of voltage safety threshold: , for emergency scoring: ; Calculate the comprehensive emergency priority by weighted comprehensive score: ; in, and are the weights of temperature and voltage, respectively, satisfying ; Priority Mapping: Low Emergency ( ): Maintain the default grouping weight; Medium Emergency ( ): Adjust the key group weights proportionally; High Emergency ( ): Forcefully increase the key group weight to the upper limit; Dynamically adjust group weights: ; in, It is a group The base weight of It is a group The emergency response factor; Ensure that the sum of all group weights is 1 to avoid over-allocation of resources and normalize the weights: .

[0041] In this embodiment, in step S43, the optimization objective and constraint conditions are defined: The objective function to minimize energy consumption is: ; in, For Channel The total loss, It is a channel The effective value of the current, It is a channel The equivalent resistance of is the switching frequency, is the switching loss energy in a single switching cycle; is the penalty term for power deviation from the rated value, It is a channel The actual output power, It is a channel Rated power; is the power deviation weight coefficient; is the total number of power supply channels in the system; The objective function to minimize the temperature rise is: ; in, For Channel Temperature rise (thermal resistance model), It is a channel The thermal resistance, It is a channel The effective cooling area; is the penalty term for the temperature rise rate; is the weight coefficient of temperature rise rate; Voltage safety constraints: ; Current balancing constraints: ; Group capacity constraints: ; Multi-objective cost function: ; in, is the weight of energy consumption and temperature rise (e.g. Indicates the focus on energy consumption optimization); According to the stack temperature Dynamic Adjustment : ; The higher the temperature, The smaller (focusing on temperature rise optimization), the optimized power supply strategy (such as duty cycle , Group ) is sent to the hardware controller.

[0042] like Figure 3 As shown, a fuel cell system domain controller FDCU is used to implement the aforementioned control method, including: A multi-dimensional grouping module, used to perform multi-dimensional grouping according to the power requirements and electrical characteristics of BOP components and generate an adaptive simplified converter topology; Condition monitoring module, used to monitor the operating parameters of BOP components in real time; Prediction module, used to predict the power demand change trend in combination with the operating condition model; A dynamic adjustment module is used to trigger group reorganization according to the prediction results and adjust the converter topology connection mode; PID control module, used to independently adjust the output of each power supply channel and dynamically optimize the grouping strategy weight.

[0043] It should be noted that the fuel cell system domain controller FDCU can implement the steps in the control method of the fuel cell system domain controller FDCU in the above embodiment, and can achieve the same technical effects. Refer to the description in the above embodiment, which will not be repeated here.

[0044] The present invention provides a fuel cell system, comprising the aforementioned FDCU, and a BOP component, a fuel cell stack and a whole vehicle high-voltage system connected to the FDCU.

[0045] The above is based on the ideal embodiment of the present invention. Through the above description, it is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention. Therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is limited by the attached claims rather than the above description, and it is intended to include all changes within the meaning and scope of the equivalent elements of the claims. Any figure mark in the claims should not be regarded as limiting the claims involved.

[0046] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A control method for a fuel cell system domain controller FDCU, characterized in that: The following steps are involved: S1. Based on the power requirements and electrical characteristics of BOP components, multi-dimensional grouping of power supply channels is performed to determine the demand boundary of each group. According to the requirements and characteristics of each group, the complex multi-port power flow is decomposed into simplified multi-port / two-port converter topologies; S2. Real-time monitoring of the operating status of BOP components, combined with historical data and operating condition models, to predict the changing trend of component power demand; S3. According to the change of power demand of BOP components, the optimal grouping scheme is selected through the transformation strategy, the grouping of power supply channels is adjusted in real time, and the converter topology connection mode is adjusted to match the power flow demand of the new grouping; S4. Control each power supply channel through PID, adjust the output voltage / current in real time, compensate for load fluctuations, and dynamically adjust the grouping strategy weight according to the system urgency to optimize the selection of the optimal power supply strategy.

2. The control method of a fuel cell system domain controller FDCU according to claim 1, characterized in that: The step S1 includes the following sub-steps: S11, establishing a multi-dimensional grouping parameter matrix, including the power level, operating voltage range, current fluctuation characteristics, response time requirements and operating frequency parameters involved; S12, based on a clustering algorithm, merging BOP components with similar parameters into the same power supply group; Objective function: Minimize the within-group squared error: ; in, is the number of groups; It is Group; It is Group center point; Grouping rules: Similarity is measured by Euclidean distance: ； in, and They are the feature vectors of two different BOP components in parameter space; and The components are and components In the Specific values ​​of the parameters; S13. Establish a dynamic demand boundary model for each power supply group, including the maximum power threshold, voltage tolerance range, and current fluctuation limit involved; Maximum power threshold , based on the maximum instantaneous power in the group and the safety margin: ; in, is the power safety factor; is the maximum instantaneous power in the group, ,in, It is a component Rated power, is the instantaneous load factor; Voltage tolerance range , based on the dynamic expansion of rated voltage and fluctuation range: ; in, Yes Group Rated voltage; is the voltage tolerance factor; Current fluctuation limitation , set the threshold based on the standard deviation of current fluctuation: ; in, Yes Group The average current of is the standard deviation of the current fluctuation within the group, , is the number of components within the group; S14. Based on the power flow path reconstruction method, a simplified multi-port / two-port converter topology is designed to optimize the power flow path. The input-output relationship of the multi-port converter is expressed as: ; in, is the number of input ports; is the number of output ports; It is The power of each input port; It is The power of each output port; It is from Input port to The transmission efficiency of each output port; The Scattering parameter matrix of the two-port converter is expressed as: ; in, and are the reflection coefficients of the input and output ports, respectively; and is the transmission coefficient, which is used to describe the power transfer characteristics between ports.

3. The control method of a fuel cell system domain controller FDCU according to claim 2, characterized in that: In the step S11, relevant parameters of all BOP components are collected, and for each BOP component, parameters are defined and quantified: the power level is divided into levels according to the rated power; the operating voltage range is based on the ratio of the voltage range width to the rated voltage; The current fluctuation characteristic is based on the ratio of the current fluctuation standard deviation to the average current; The response time requirement is a normalized response time; The operating frequency parameter is based on the offset ratio between the operating frequency and the base frequency; For the multidimensional grouping parameter matrix, the weight vector is multiplied by the parameter matrix to obtain a weighted matrix : ; in, is a multidimensional grouping parameter matrix; is the transpose of the weight vector; is the power level; It is a component The weight vector of .

4. The control method of a fuel cell system domain controller FDCU according to claim 1, characterized in that: The step S2 includes the following sub-steps: S21, real-time monitoring of dynamic power requirements, electrical parameter deviations and operation mode switching signals of BOP components; S22. Predict the power demand change trend of BOP components using the operating condition model; Operating model combining physical equations and data-driven models: ; in, is the weight coefficient; It is a predicted value based on the physical model; is the machine learning output; S23, predicting the triggering threshold of grouping adjustment through the grouping decision model trained with historical data; The group decision model is quickly deployed through logistic regression: ; in, It's at the time , the probability of triggering group adjustment, range ; is the intercept term of the model; It is Features The regression coefficient of .

5. The control method of a fuel cell system domain controller FDCU according to claim 1, characterized in that: The step S3 includes the following sub-steps: S31, when the power or electrical parameters of the monitored BOP components exceed the current grouping threshold, dynamic grouping reorganization is triggered; Model output trigger probability, set dynamic threshold based on historical data distribution : ; in, is the recent average trigger probability; is the standard deviation; is the sensitivity coefficient; S32, selecting the optimal grouping scheme through a decision tree algorithm and reallocating power supply channels; The decision tree model selects split features and thresholds to minimize the Gini impurity of child nodes: ; in, is the total number of categories in the dataset; It is the first A subset of samples of each category; is the current data set to be segmented; S33, switch the parallel / series mode of the converter topology, use soft switching technology to reduce transient impact, and calibrate the topology parameters of the reorganized power supply channel to ensure that the power flow matches the demand.

6. The control method of a fuel cell system domain controller FDCU according to claim 5, characterized in that: In the step S33, for topology switching, when the original parallel power supply channel is switched to the series mode: ; in, is the output voltage of the new topology after switching; and is the output voltage of the original parallel power supply channel; is the output current of the new topology after switching; and is the output current of the original parallel power supply channel; Add zero voltage switching condition during switching: ; in, is the resonant inductance value, used to achieve ZVS; is the resonant circuit peak current; is the output capacitance of the switch tube; is the drain-source voltage of the switch tube; Adjust the converter inductor / capacitor values ​​so that the output matches the power flow requirements of the new group: ; in, is the calibrated inductance value; is the converter input voltage; is the duty cycle; is the allowable current ripple; is the switching frequency; Check whether the reorganized power flow satisfies: ; in, is the actual measured power; is the target power; is the allowable relative error.

7. The control method of a fuel cell system domain controller FDCU according to claim 1, characterized in that: The step S4 includes the following sub-steps: S41. Design an independent PID controller for each power supply channel to adjust the output voltage / current in real time to compensate for load fluctuations; Continuous domain PID control law: ; in, is the control output; is the error signal, or ,in, is the voltage setting value, is the actual output voltage, is the current setting value, is the actual output current; is the cumulative amount of error; is the rate of change of the error; , and are the proportional, integral and differential coefficients respectively; is the current time, indicating the real-time running time of the controller; is the integral variable, which represents the time integral process from the initial moment to the current moment; S42, setting strategy priorities according to the temperature and voltage safety thresholds of the battery stack, and dynamically adjusting the grouping weights; S43, selecting a power supply strategy by optimizing a cost function, wherein the objective function includes minimum energy consumption or minimum temperature rise; The objective function to minimize energy consumption is: ; in, For Channel The total loss, It is a channel The effective value of the current, It is a channel The equivalent resistance of is the switching frequency, is the switching loss energy in a single switching cycle; is the penalty term for power deviation from the rated value, It is a channel The actual output power, It is a channel Rated power; is the power deviation weight coefficient; is the total number of power supply channels in the system; The objective function to minimize the temperature rise is: ; in, For Channel The temperature rise, It is a channel The thermal resistance, It is a channel The effective cooling area; is the penalty term for the temperature rise rate; is the weight coefficient of temperature rise rate; Multi-objective cost function: ; in, is the weight of energy consumption and temperature rise.

8. The control method of a fuel cell system domain controller FDCU according to claim 7, characterized in that: In the step S42, the stack temperature is defined as and the voltage is safe Threshold: Safety range of stack temperature threshold: , for emergency scoring: ; Safety range of voltage safety threshold: , for emergency scoring: ; Calculate the comprehensive emergency priority by weighted comprehensive score: ; in, and are the weights of temperature and voltage respectively; Priority mapping: Low urgency: maintain the default group weight; Medium urgency: adjust the weight of the key group proportionally; High urgency: forcibly increase the weight of the key group to the upper limit; Dynamically adjust group weights: ; in, It is a group The base weight of It is a group emergency response factor.

9. A fuel cell system domain controller FDCU, used to implement the control method according to any one of claims 1 to 8, characterized in that: include: A multi-dimensional grouping module, used to perform multi-dimensional grouping according to the power requirements and electrical characteristics of BOP components and generate an adaptive simplified converter topology; Condition monitoring module, used to monitor the operating parameters of BOP components in real time; Prediction module, used to predict the power demand change trend in combination with the operating condition model; A dynamic adjustment module is used to trigger group reorganization according to the prediction results and adjust the converter topology connection mode; PID control module, used to independently adjust the output of each power supply channel and dynamically optimize the grouping strategy weight.

10. A fuel cell system, characterized in that: It comprises the FDCU as claimed in claim 9, and BOP components, a fuel cell stack and a whole vehicle high voltage system connected to the FDCU.

Images