Optimization method of power management system of triboelectric nanogenerator based on genetic algorithm

Abstract

The present invention discloses a method for optimizing a friction nanogenerator power management system based on a genetic algorithm. The method simulates different TENG forms to obtain corresponding characterizations, and uses two main indicators, power curve and equivalent capacitance interval, for characterization matching, so that the simulation results have a strong consistency with the measured data. Then, the NSGA‑Ⅱ multi-objective non-dominated sorting genetic algorithm is used, which can simultaneously perform multi-objective optimization on the output efficiency and matching coefficient, realize the optimized output of the same circuit under different parameter combinations, and can perform specific optimization comparisons of different structures for the special power device TENG. In the entire optimization design process, the parameter selection of circuit components can be quickly realized, and the design of special power management systems under different forms of loads and different forms of TENG can be accelerated, providing certain support for the implementation of more applications of TENG.

Description

Technical Field

[0001] The present invention belongs to the technical field of system design optimization, and relates to a power management system of a friction nanogenerator, and in particular to an optimization method of a power management system of a friction nanogenerator based on a genetic algorithm. Background Art

[0002] The Internet of Things (IoT) technology has become an indispensable part of our lives, and sensors, as the end of the IoT and interconnected systems, play a vital role. However, the problem of how to provide working energy for a large number of sensors in different environments has become one of the main reasons restricting the further development of the IoT technology.

[0003] Triboelectric Nanogenerator (TENG) is an emerging energy supply device based on the combination of triboelectric effect and electrostatic induction, which has incomparable advantages. Triboelectric nanogenerator TENG is mainly installed in an environment rich in mechanical energy. In theory, it can collect energy for an unlimited number of times to power sensors, and can be used in a complementary manner with traditional batteries, greatly extending the single use time of sensors, and to a certain extent solving the problem of long-term energy supply required for network nodes in wireless sensor networks.

[0004] There are two main ways to use TENG in sensors: (1) TENG as an energy collection module; (2) TENG as a self-powered sensor that does not require energy. Whether it is used as an energy collection module or a self-powered sensor module, at least two parts are required: the TENG itself and its peripheral circuit. In the prior art, the research focus on TENG and its peripheral circuits is still on TENGs with simple structures and simple motion modes and circuits with resistor-capacitor loads as the main loads. The research method is mainly to use analytical models and circuit theory for analysis. However, in practical applications, the TENG structure, motion mode or circuit structure are often more complicated. At this time, it is difficult to obtain a better analytical solution when using analytical models for analysis. In addition, for TENGs of different forms, the output is not constant, and different TENG forms often have a great impact on the corresponding output conditions, thereby affecting the design of the system.

[0005] TENG is a special energy harvesting device with high voltage and low current output. Its output frequency is greatly affected by environmental changes. Traditional power management circuits are not directly applicable to this device. Therefore, it is necessary to design a dedicated power management system based on the specific structure and motion state of TENG. At present, many teams have developed dedicated power management circuits for TENG. [1]Conduct research, including trying to use finite element models to conduct specific research on TENG morphological output analysis. Using computer simulation optimization technology to optimize the design of TENG power management systems can greatly reduce the cost of trial and error, reduce the design cycle of the overall TENG power management circuit, and greatly improve design efficiency. Summary of the invention

[0006] In view of the shortcomings of the prior art, the present invention proposes a friction nanogenerator power management system optimization method based on genetic algorithm. According to the timing output of the selected TENG, the corresponding power management circuit is determined, and the parameters of the power management circuit components determined by the NSGA-Ⅱ genetic algorithm are optimized to avoid output deviation caused by the power management system.

[0007] The optimization method of the friction nanogenerator power management system based on genetic algorithm specifically includes the following steps:

[0008] Step 1: Finite element model simulation and output matching

[0009] The basic characterization parameters of TENG include the output open-circuit voltage and the equivalent capacitance. The output open-circuit voltage and the equivalent capacitance under the basic motion state can realize the basic output characterization of TENG in various forms. Finite element model simulation includes three steps: static model drawing, simplification and dynamic characterization value output. The simulation output of the finite element model is compared with the measured value to obtain the simulation output that matches the actual output.

[0010] s1.1. Draw two-dimensional and three-dimensional finite element static general simulation models based on the structure and material parameters of TENG, and measure the actual output power curve and capacitance change of TENG.

[0011] s1.2. According to the actual motion state of TENG during use and the differences in capacitance and voltage output, the model obtained in s1.1 is simplified to obtain a capacitance model for capacitance output and a voltage model for voltage output, respectively, to reduce the difficulty of simulation mesh drawing and boundary condition setting.

[0012] s1.3. Simulate the motion process of TENG, use the pure open-circuit voltage value in the voltage model as the output open-circuit voltage, and use the inter-electrode capacitance value in the capacitance model as the output capacitance. Through interpolation and replication, the complete multi-cycle equivalent capacitance output and voltage output time-varying sequence are obtained.

[0013] s1.4. Draw the power output curve and capacitance output range of the TENG finite element model according to the output time-varying sequence obtained in s1.3, and compare them with the measured power output curve and capacitance output range obtained in s1.1. When the simulated output does not match the measured data, return to s1.2, adjust the parameters in the capacitance model and voltage model, repeat s1.3 and s1.4 until the output matches, and save the TENG characterization timing output result at this time.

[0014] Preferably, in s1.3, the movement process of TENG is simulated by deformation geometry or parametric scanning.

[0015] Preferably, in s1.4, the output change of the capacitance model is first adjusted to be consistent with the measured capacitance change, and then the parameters in the model affected by the environment are adjusted until the output power curve of the model matches the measured output power curve.

[0016] Step 2: Optimize the design of power management system

[0017] The structural parameters, material parameters of TENG and the component parameters in the dedicated power management circuit are input into the NSGA-Ⅱ genetic algorithm together. High output efficiency and low matching point are taken as the optimization goals. The dedicated power management circuit and its component parameters that meet the design requirements are obtained through iteration.

[0018] Preferably, the structural parameters and material parameters of TENG include electrode rotor speed, electrode effective area, and dielectric constant and thickness of friction material.

[0019] s2.1. Input the structural parameters, material parameters, and component parameters of the TENG and the dedicated power management circuit into NSGA-Ⅱ to initialize the population. The component parameters of the power management circuit are set as variable parameters.

[0020] s2.2. Through the finite element simulation and input matching method of step one, the capacitor voltage time series output by the TENG finite element model is obtained, and then the capacitor voltage time series and the component parameters set in s2.1 are input into the dedicated power management circuit model to simulate the output of the load voltage and current.

[0021] s2.3. Calculate the output efficiency and matching coefficient of the dedicated power management circuit under different component parameters, and complete the individual fitness calculation of the NSGA-Ⅱ genetic algorithm. Then set two optimization goals of high output efficiency and low matching coefficient, and use the NSGA-Ⅱ genetic algorithm to screen, copy, cross, and mutate different parameter combinations to obtain new parameter combinations. After sorting, screening, and N iterative optimizations, output the parameter combination, select it according to actual use requirements, and obtain the optimized dedicated power management circuit.

[0022] Preferably, the dedicated power management circuit is a thyristor capacitor two-level charging power management circuit.

[0023] Preferably, in each iterative optimization process, 8 groups of different parameter combinations are generated.

[0024] Preferably, the number of iterations is set to 200.

[0025] The present invention has the following beneficial effects:

[0026] A method for optimizing the circuit design of a power management system for a friction nanogenerator based on computer simulation technology is proposed. By simulating different TENG forms, the corresponding characterization is obtained, and the power curve and equivalent capacitance range are used as two main indicators for characterization matching, so that the simulation results are highly consistent with the measured data. Then, the NSGA-Ⅱ multi-objective non-dominated sorting genetic algorithm is used to simultaneously perform multi-objective optimization on the output efficiency and matching coefficient, and achieve the optimized output of the same circuit under different parameter combinations, and can perform specific optimization comparisons of different structures for the special power device TENG. In the entire optimization design process, the parameter selection of circuit components can be quickly realized, and the design of special power management systems under different load forms and different forms of TENG can be accelerated, providing certain support for the implementation of more applications of TENG. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 It is a flow chart of finite element model simulation and output matching;

[0028] Figure 2 It is a flow chart of power management system optimization design;

[0029] Figure 3 is a schematic diagram of a dedicated power management circuit used in the embodiment;

[0030] Figure 4 is a schematic diagram of the capacitor model structure obtained after simplification in the embodiment;

[0031] Figure 5 is a schematic diagram of a simplified voltage model structure in the embodiment;

[0032] Figure 6 It is the arrangement of the single circuit parameter combination group after simulation optimization in the embodiment. DETAILED DESCRIPTION

[0033] The present invention will be further explained below with reference to the accompanying drawings;

[0034] The optimization method of the friction nanogenerator power management system based on genetic algorithm specifically includes the following steps:

[0035] Step 1: Finite element model simulation and output matching

[0036] The premise of optimizing the design based on computer simulation technology is to match the simulation results with the measured data. The basic characterization parameters of TENG include the output open-circuit voltage and the equivalent capacitance. By obtaining the basic output open-circuit voltage and the equivalent capacitance, the basic output characterization of various forms of TENG can be achieved. Figure 1 As shown, the finite element model simulation includes three steps: static model drawing, simplification and dynamic characterization value output.

[0037] The time-varying array files of equivalent capacitance and open-circuit voltage are obtained through finite element simulation. First, the capacitance change of TENG is measured and simulated to ensure that the capacitance output change range during simulation is basically consistent with the capacitance output obtained by actual measurement. Then, the power curve of the simulation output is compared with the power curve obtained by actual test with an oscilloscope. After continuously adjusting the parameters that are greatly affected by the environment, such as the surface charge density, and obtaining relatively consistent results, it can be considered that the obtained capacitance and voltage time-varying array files can fully characterize the actual TENG output state, and the time-varying sequence files of CTENG and Voc are derived.

[0038] Step 2: Optimize the design of power management system

[0039] like Figure 2 As shown in the figure, the structural parameters and material parameters of TENG and the component parameters in the dedicated power management circuit are input into the NSGA-Ⅱ genetic algorithm, and high output efficiency and low matching point are used as optimization targets. The dedicated power management circuit and its component parameters that meet the design requirements are obtained through iteration. Not only can the TENG dedicated power management circuits with the same function be compared horizontally, but also the single circuit optimized by the genetic algorithm iteration can be compared vertically, and finally the component parameters of the optimal circuit for the given optimization target can be selected.

[0040] s2.1. Input the structural parameters, material parameters, and component parameters of the dedicated power management circuit of TENG into NSGA-Ⅱ to initialize the population. The component parameters of the power management circuit are set as variable parameters. In order to achieve a higher efficiency DC output, the power management circuit structure used in this embodiment is as follows: Figure 3 As shown, it is a thyristor capacitor two-stage charging power management circuit with good charging effect. [2], using the thyristor part to control the multi-level capacitor charging energy and provide DC energy for the subsequent load R1. The Voc and CTENG on the left represent the matched TENG equivalent input. The parameters that need to be optimized in this circuit include Cin, Cout, D5, D6, L1 and the component values ​​of the SCR thyristor.

[0041] s2.2. Through the finite element simulation and input matching method of step one, the capacitor voltage time series output by the TENG finite element model is obtained, and then the capacitor voltage time series and the component parameters set in s2.1 are input into the dedicated power management circuit model to simulate the output of the load voltage and current.

[0042] In this embodiment, the finite element simulation model of the open circuit voltage of the rotary TENG is selected. Specifically, the friction material of the TENG is FEP, and the turntable is a stator single-sided double electrode made using the PCB manufacturing process. Due to the different simulation forms of capacitance and voltage, in the process of model simplification, the thickness of the rotor brass blades and the thickness of the stator brass double electrodes are mainly simplified to varying degrees. The simplified capacitance model and voltage model are shown in Figure 1. Figure 4 , Figure 5 Using a simplified model can reduce the difficulty of drawing the simulation mesh and setting the boundary conditions.

[0043] s2.3. Calculate the output efficiency and matching coefficient of the dedicated power management circuit under different component parameters, and complete the individual fitness calculation of the NSGA-Ⅱ genetic algorithm. Select higher output efficiency and lower matching points as the two optimization goals, where lowering the matching point is represented by the matching coefficient, that is, the proportion of the decrease in the optimal matching impedance. Use the NSGA-Ⅱ genetic algorithm to screen, copy, cross, and mutate different parameter combinations to obtain new parameter combinations. After sorting and screening, repeat the iterative process 200 times, and generate 8 different parameter combinations in each iteration. After N iterations, the algorithm converges, and the output optimized parameter combination is compared with the randomly generated parameter combination. Figure 6 As shown, the star-shaped individual group represents the output state of the randomly generated individual, and the triangle group represents the output state of the optimized individual group. It can be seen that the optimized group is relatively close to the coordinate axis, which can ensure the basic requirements of at least one of the two indicators of high efficiency or low matching coefficient. By recording the optimization results of each generation, the specific differences in the efficiency and output matching point drop of different circuit parameters under different iteration states can be compared. Similarly, under the optimal state, the circuit efficiency and output matching point drop of different parameter circuits can be compared more intuitively, which is of great benefit to the selection of different TENG-specific power management circuits.

[0044] In addition to the above contents, this method can also perform simulation characterization on more complex RLC loads, obtain different optimal power supply and management structure design methods based on different load characterization results, and optimize the parameters of the same-type TENG for a given specific power management circuit under the complete characterization of different structural parameters.

[0045] [1]Xc A, Wei TB, Yu SA, et al. Power management and effective energy storage of pulsed output from triboelectric nanogenerator-ScienceDirect[J]. Nano Energy, 2019, 61: 517-532.

[0046] [2]Harmon W,Bamgboje D,Guo H,et al.Self-driven Power ManagementSystem for Triboelectric Nanogenerators[J].Nano Energy,2020,71(8):104642.

Claims

1. A method for optimizing a power management system of a triboelectric nanogenerator based on a genetic algorithm, characterized in that: The method specifically comprises the following steps: Step 1: Finite element model simulation and output matching A static universal simulation model is drawn according to the structure and material parameters of the friction nanogenerator, and the actual output power and capacitance of the friction nanogenerator under different motion states are measured; the static universal simulation model is simplified to obtain a capacitance model and a voltage model, and the capacitance model parameters and the voltage model parameters are continuously adjusted until the output power of the static universal simulation model is consistent with the actual output power obtained by measurement, and the timing output results of the friction nanogenerator characterization are saved, and the complete multi-cycle equivalent capacitance output and voltage output time series are obtained through interpolation and replication; Step 2: Optimize the design of power management system s2.

1. Input the structural parameters, material parameters of the friction nanogenerator, and the component parameters in the dedicated power management circuit into the NSGA-Ⅱ genetic algorithm to initialize the population; the component parameters of the power management circuit are set as variable parameters; s2.

2. By using the finite element simulation and output matching method in step 1, the capacitor voltage time series output by the finite element model of the friction nanogenerator is obtained, and then the capacitor voltage time series and the component parameters set in s2.1 are input into the dedicated power management circuit model to simulate the output of the load voltage and current; s2.

3. Calculate the output efficiency and matching coefficient of the dedicated power management circuit under different component parameters, and complete the individual fitness calculation of the NSGA-Ⅱ genetic algorithm; then set two optimization goals of high output efficiency and low matching coefficient, and use the NSGA-Ⅱ genetic algorithm to screen, copy, cross, and mutate different parameter combinations to obtain new parameter combinations. After sorting, screening, and N iterative optimizations, output the parameter combination to obtain the optimized dedicated power management circuit.

2. The method for optimizing the power management system of a triboelectric nanogenerator based on a genetic algorithm as claimed in claim 1, characterized in that: The step 1 specifically includes: s1.

1. Draw a two-dimensional and three-dimensional finite element static general simulation model based on the structure and material parameters of the friction nanogenerator, and measure the actual output power curve and capacitance change of the friction nanogenerator; s1.

2. According to the actual motion state of the friction nanogenerator during use and the differences in capacitance and voltage output, the model obtained in s1.1 is simplified to obtain a capacitance model for capacitance output and a voltage model for voltage output respectively; s1.

3. Simulate the motion process of the friction nanogenerator, use the pure open-circuit voltage value in the voltage model as the output open-circuit voltage, and use the inter-electrode capacitance value in the capacitance model as the output capacitance; obtain the complete multi-cycle equivalent capacitance output and voltage output time-varying sequence through interpolation and replication; S1.

4. Draw the power output curve of the friction nanogenerator finite element model according to the output time-varying sequence obtained in S1.3, and compare it with the measured power output curve obtained in S1.

1. When the simulated output does not match the measured data, adjust the parameters in the capacitance model and the voltage model until they match, and save the friction nanogenerator characterization timing output result at this time.

3. The method for optimizing the power management system of a triboelectric nanogenerator based on a genetic algorithm as claimed in claim 2, characterized in that: In s1.3, the motion process of the friction nanogenerator is simulated by deformation geometry or parametric scanning.

4. The method for optimizing the power management system of a triboelectric nanogenerator based on a genetic algorithm as claimed in claim 2, characterized in that: In s1.4, the output change of the capacitor model is first adjusted to be consistent with the measured capacitance change, and then the parameters in the model affected by the environment are adjusted until the output power curve of the model matches the measured output power curve.

5. The method for optimizing the power management system of a triboelectric nanogenerator based on a genetic algorithm as claimed in claim 1, characterized in that: The structural parameters and material parameters of the friction nanogenerator include the electrode rotor speed, electrode effective area, and the dielectric constant and thickness of the friction material.

6. The method for optimizing a friction nanogenerator power management system based on a genetic algorithm as claimed in claim 1, characterized in that: The dedicated power management circuit is a thyristor capacitor secondary charging power management circuit.

7. The method for optimizing a power management system of a triboelectric nanogenerator based on a genetic algorithm as claimed in claim 1, characterized in that: In each iterative optimization process, 8 sets of different parameter combinations are generated.

8. The method for optimizing a triboelectric nanogenerator power management system based on a genetic algorithm as claimed in claim 1 or 7, characterized in that: Set the number of iterations to 200.

Images