A method for debugging a power conversion system (PCS)

Abstract

The application provides a debugging method of an energy storage converter PCS, comprising the following steps: according to a switching characteristic-temperature mapping relationship of a power semiconductor device, combining a real-time temperature detection value obtained, dynamically adjusting a turn-on time and a turn-off time of an IGBT or a MOSFET, and optimizing a switching timing sequence of the power semiconductor device to adapt to working states under different temperatures; through real-time sampling and analysis on output voltage and current waveforms of the energy storage converter PCS, calculating a total harmonic distortion rate and a power factor index of the output waveforms, evaluating output waveform quality, and if the waveform quality index does not reach a preset requirement, further optimizing switching timing sequence parameters of the power semiconductor device.

Description

Technical Field

[0001] The present invention relates to the field of information technology, and in particular to a debugging method for a PCS (Peripheral Control System) of an energy storage converter. Background Art

[0002] During the commissioning of energy storage converter PCSs, the impact of power semiconductor device switching characteristics on the PCS's dynamic response is a complex and critical technical issue. As the core switching devices in PCSs, IGBTs or MOSFETs face significant differences in their turn-on and turn-off times, as well as parameter drift caused by temperature variations. These differences directly impact the quality of the PCS's output waveform and conversion efficiency. However, the typical parameters listed in datasheets for different power semiconductor models often differ significantly from their dynamic characteristics under actual application conditions. This makes it difficult for PCS control algorithms to accurately estimate switching delays and compensate for dead-time, leading to a series of issues such as output current distortion and DC bus voltage fluctuations. Furthermore, the ambient temperature range within which PCSs operate varies widely, and the dynamic characteristics of power semiconductors are highly sensitive to junction temperature. Even if margins are factored in during selection, fluctuations in junction temperature can still lead to unpredictable variations in switching characteristics during actual operation. This is especially true for high-power devices, where uneven temperature gradients within the device exacerbate this uncertainty. Therefore, how to monitor the dynamic characteristics of power semiconductor devices in real time on a PCS machine and adaptively optimize switching timing and dead-zone compensation accordingly to ensure stable output of high-quality electrical energy under various operating conditions is a technical challenge that needs to be overcome urgently. Summary of the Invention

[0003] The present invention provides a debugging method for a PCS energy storage converter, which mainly includes:

[0004] Obtain the type parameters of the power semiconductor devices in the energy storage converter PCS, including the model, rated voltage, and rated current of the IGBT or MOSFET. Based on the parameters of the power semiconductor devices, determine the turn-on time and turn-off time characteristic curves of the power semiconductor devices at different temperatures, and obtain the mapping relationship between the switching characteristics of the power semiconductor devices and temperature;

[0005] A temperature sensor is used to detect the operating temperature of the power semiconductor devices in the energy storage converter PCS in real time. The detected temperature value is compared with the preset temperature threshold. If the detected temperature exceeds the threshold, it is determined that the power semiconductor device is operating in a high temperature state and the switching timing needs to be adjusted to adapt to the temperature change.

[0006] Based on the mapping relationship between the switching characteristics and temperature of power semiconductor devices, combined with the real-time temperature detection value, the on-time and off-time of IGBT or MOSFET are dynamically adjusted to optimize the switching timing of power semiconductor devices to adapt to the working conditions at different temperatures.

[0007] By sampling and analyzing the output voltage and current waveforms of the energy storage converter PCS in real time, the total harmonic distortion rate and power factor of the output waveform are calculated, and the output waveform quality is evaluated. If the waveform quality indicators do not meet the preset requirements, the switching timing parameters of the power semiconductor devices are further optimized;

[0008] A power analyzer is used to measure the input power and output power of the energy storage converter PCS, calculate the operating efficiency of the energy storage converter PCS, and compare the measured efficiency value with a preset efficiency threshold. If the efficiency is lower than the threshold, the switching timing parameters of the power semiconductor device are adjusted to reduce switching losses.

[0009] For the different operating conditions of the energy storage converter PCS, including light load, heavy load, and overload, a corresponding optimal switching timing parameter table for power semiconductor devices is established. The switching timing parameter table clearly lists the optimal turn-on and turn-off times under various load conditions, including different turn-on and turn-off times corresponding to light load, heavy load, and overload conditions. During the operation of the energy storage converter PCS, the corresponding switching timing parameters are switched in real time according to the load conditions;

[0010] Based on the switching characteristics, operating temperature, output waveform quality and working efficiency of the power device of the energy storage converter PCS, a multi-objective optimization model is constructed. The particle swarm optimization algorithm is used to globally optimize the switching timing parameters. The switching timing control strategy with the best comprehensive performance is obtained to guide the debugging and operation of the PCS.

[0011] The technical solution provided by the embodiment of the present invention may have the following beneficial effects:

[0012] The present invention discloses a debugging method for a PCS (power storage converter). The method obtains power device parameters and establishes a mapping relationship between switch characteristics and temperature, monitors device temperature in real time and compares it with a threshold value, and dynamically adjusts the switch timing according to the temperature. At the same time, the switch parameters are further optimized by analyzing the output waveform quality and efficiency. An optimal switch timing parameter table is established for different load conditions to achieve load adaptive control. Finally, a multi-objective optimization model is constructed, and a particle swarm algorithm is used for global optimization to obtain a switch timing strategy with the best overall performance. The present invention can effectively improve the working performance of the PCS under different temperature and load conditions, reduce switching losses, and improve system efficiency and output waveform quality, and has important engineering application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 The present invention is a flowchart of a method for debugging an energy storage converter PCS.

[0014] Figure 2 A schematic diagram of a debugging method for an energy storage converter PCS according to the present invention.

[0015] Figure 3 Another schematic diagram of a debugging method of a power storage converter PCS. DETAILED DESCRIPTION

[0016] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application will be described in detail below with reference to the drawings and specific embodiments.

[0017] As Figure 1-3 , the debugging method of the power storage converter PCS can specifically include:

[0018] S101, obtain the type parameters of the power semiconductor device in the power storage converter PCS, including the model, rated voltage and rated current of IGBT or MOSFET, determine the turn-on time and turn-off time characteristic curves of the power semiconductor device at different temperatures according to the parameters of the power semiconductor device, and obtain the switching characteristic and temperature mapping relationship of the power semiconductor device.

[0019] Obtain the model, rated voltage and rated current parameters of the power semiconductor device in the power storage converter PCS, establish a static parameter database of the power semiconductor device according to the parameters; perform temperature scanning test on the power semiconductor device to obtain a mapping relationship model of switching characteristics and temperature; calculate the switching loss and conduction loss of the PCS under different working states according to the static parameter database and the mapping relationship model; import the switching loss and conduction loss data into thermal simulation software to analyze the junction temperature distribution of the power semiconductor device; if the junction temperature distribution exceeds the maximum junction temperature specified by the device, use a particle swarm optimization algorithm to optimize the topology structure of the PCS; build a PCS control model according to the optimized topology structure; perform iterative calculation through the PCS control model to obtain the optimal operating parameters of the PCS.

[0020] Specifically, the model, rated voltage, and rated current of the power semiconductor devices in the energy storage converter PCS were obtained. A database was established based on the static parameters of the power semiconductor devices. The basic characteristic parameters of different types of power semiconductor devices were obtained by querying the database. A temperature sweep test was performed using a Keysight B1506A power device analyzer, measuring the turn-on and turn-off times of the power semiconductor devices in 10°C intervals over the temperature range of -40°C to 150°C. The test data was recorded and curve fit was performed using the cftool toolbox in MATLAB software, selecting the polynomial fitting method to obtain a model mapping the switching characteristics and temperature. Based on the obtained power semiconductor device parameters and the switching characteristic model, the switching and conduction losses of the PCS under different operating conditions were calculated using PLECS simulation software. The calculated loss data was imported into ANSYS Icepak thermal simulation software, and heat dissipation conditions and boundary parameters were set to analyze the junction temperature distribution of the power semiconductor devices. The junction temperature was determined to determine whether it exceeded the maximum junction temperature specified in the device data sheet, thereby confirming whether the device met the PCS heat dissipation requirements. Based on the switching characteristic model and loss calculation results of power semiconductor devices, a particle swarm optimization algorithm was used to optimize the PCS topology and adjust the switching frequency and current stress. A PCS control model was built using MATLAB / Simulink to optimize the closed-loop control strategy. Through iterative calculations, the optimal operating parameters of the PCS were obtained, improving the conversion efficiency of the energy storage converter. The parameters of the power semiconductor devices in the energy storage converter PCS were obtained from a database. For example, the IGBT model was FF300R12KT4, with a rated voltage of 1200V and a rated current of 300A. Temperature sweep testing was performed using a Keysight B1506A power device analyzer, measuring the turn-on and turn-off times at 10°C intervals over a temperature range of -40°C to 150°C. For example, at 25°C, the turn-on time was measured to be 200ns and the turn-off time to be 300ns. Using polynomial fitting, the MATLAB cftool toolbox derived a mapping model for switching characteristics and temperature: Ton(T) = 0.01T² + 0.5T + 180°C, Toff(T) = 0.015T² + 0.8T + 270°C, where T is temperature (°C) and the on-time, Ton, and off-time, Toff, are expressed in nanoseconds. Based on this model, the PLECS simulation software calculates the PCS losses under different operating conditions. For example, at a 50kHz switching frequency, a 400V DC bus voltage, and a 100A load current, the switching losses are 150W and the conduction losses are 200W. Using ANSYS Icepak thermal simulation software, cooling conditions are set, such as a heatsink thermal resistance of 0.1°C / W, and boundary parameters, such as an ambient temperature of 40°C. The analysis yielded an IGBT junction temperature of 125°C, lower than the datasheet's maximum junction temperature of 150°C and meeting the PCS's cooling requirements.The particle swarm optimization algorithm optimizes the PCS topology, the number of particles is set to 50, and the maximum number of iterations is 100. By adjusting the switching frequency (range 30-100 kHz) and current stress, its range is 50-150 A, the optimal switching frequency is 60 kHz, and the current stress is 120 A. The MATLAB / Simulink builds a PCS control model to realize the automatic adjustment of the PI controller parameters, such as the proportional coefficient Kp from 0.1 to 10, and the integral coefficient Ki from 1 to 100 for scanning, and the optimal parameters Kp = 5 and Ki = 50 are obtained. After 10 iterations, the optimal operating parameters of the PCS are obtained: switching frequency 60 kHz, DC bus voltage 420 V, and current stress 120 A, and the energy storage converter conversion efficiency reaches 98.5%.

[0021] In S102, a temperature sensor is used to detect the working temperature of the power semiconductor device in the energy storage converter PCS in real time, and the detected temperature value is compared with the preset temperature threshold. If the detected temperature exceeds the threshold, it is judged that the power semiconductor device is working in a high temperature state, and the switching time sequence needs to be adjusted to adapt to the temperature change.

[0022] A thermocouple temperature sensor is installed on the surface of the power semiconductor device, and temperature data is collected in real time by a data acquisition card at a preset sampling frequency. The collected temperature data is converted into a digital signal and input into the PCS controller. According to the preset multi-level temperature threshold in the PCS controller, the comparator circuit compares the detected temperature with the preset temperature threshold in real time. If the detected temperature exceeds the first warning threshold, the temperature compensation in the PCS controller is triggered. The temperature compensation looks up the corresponding switching frequency in the preset temperature and frequency correspondence table according to the current temperature, and adjusts the duty cycle of the PWM wave according to the temperature and duty cycle correspondence table. If the detected temperature continues to rise and exceeds the second warning threshold, the PCS controller further reduces the switching frequency and current stress of the power semiconductor device. If the temperature continues to rise, the preset multi-level cooling system is started, including air cooling and water cooling systems, until the temperature decreases or reaches the over-temperature protection threshold.

[0023] Specifically, thermocouple temperature sensors are installed on the surface of power semiconductor devices. A data acquisition card collects temperature data in real time at a sampling frequency of 1kHz. The collected temperature values ​​are converted into digital signals and input into the PCS controller. Temperature thresholds are preset in the PCS controller. Based on the operating characteristics of the power semiconductor devices and the design of the cooling system, multiple temperature thresholds are set, such as 85°C as the first warning threshold, 100°C as the second warning threshold, and 120°C as the overtemperature protection threshold. A multi-stage cooling system, including air cooling and water cooling, is also designed, with preset activation temperature thresholds for each level. A comparator circuit compares the detected temperature with the preset temperature thresholds in real time. If the detected temperature exceeds the first warning threshold, the temperature compensation algorithm in the PCS controller is triggered. This algorithm uses a lookup table approach to search for the appropriate switching frequency based on the current temperature in a preset temperature-frequency mapping table. It also adjusts the duty cycle of the PWM wave according to the temperature-duty cycle mapping table. For example, when the temperature rises from 85°C to 90°C, the switching frequency decreases from 50kHz to 45kHz, and the duty cycle decreases from 0.8 to 0.75. If the detected temperature continues to rise and exceeds the second-level warning threshold, the PCS controller further reduces the switching frequency and current stress of the power semiconductor devices. Specifically, the switching frequency is linearly reduced, for example, from 45kHz to 40kHz; the current stress is simultaneously reduced by 20%, for example, from 100A to 80A. Furthermore, forced air cooling is activated to improve heat dissipation efficiency. If the temperature continues to rise, the water cooling system is gradually activated until the temperature drops or reaches the overtemperature protection threshold. In the energy storage converter PCS, thermocouple temperature sensors, such as K-type, are mounted on the surface of the IGBT module. A 16-bit ADC data acquisition card collects temperature data in real time at a sampling rate of 1kHz. The acquired analog signal is converted into a digital signal and input into the PCS controller. The PCS controller is based on a DSP chip, such as TI's TMS320F28335, with preset temperature thresholds of 85°C (first-level warning), 100°C (second-level warning), and 120°C (overtemperature protection). The cooling system includes two levels: air cooling, with a startup temperature of 85°C, and water cooling, with a startup temperature of 100°C. The comparator circuit uses an LM339 chip to compare the detected temperature with a preset threshold in real time. When the temperature exceeds 85°C, a temperature compensation algorithm is triggered. This algorithm, implemented in the DSP, uses a 256×2 lookup table to store the temperature-frequency and temperature-duty-cycle relationships. For example, when the temperature rises from 85°C to 90°C, the table indicates a reduction in switching frequency from 50kHz to 45kHz and a reduction in duty cycle from 0.8 to 0.75. The IGBT driver circuit adjusts the PWM waveform accordingly. If the temperature continues to rise to 100°C, the controller linearly reduces the switching frequency to 40kHz and simultaneously adjusts the PWM waveform to reduce the current stress from 100A to 80A. At this point, the controller outputs a signal through the GPIO port to activate a DC fan rated at 200W to increase heat dissipation.If the temperature continues to rise, the water cooling system is activated at 110°C, with a water pump flow rate of 10L / min. The DSP continuously monitors temperature changes, recording temperature data every 100ms and storing it in the onboard EEPROM for subsequent analysis. If the temperature reaches 120°C, overtemperature protection is triggered, and the controller shuts down the IGBT gate drive, cutting off the main circuit current.

[0024] S103. According to the mapping relationship between the switching characteristics and temperature of the power semiconductor device and in combination with the acquired real-time temperature detection value, the on-time and off-time of the IGBT or MOSFET are dynamically adjusted to optimize the switching timing of the power semiconductor device to adapt to the working conditions at different temperatures.

[0025] A K-type thermocouple temperature sensor is used to collect temperature data of power semiconductor devices in real time. Based on a pre-established database of mapping relationships between the switching characteristics of power semiconductor devices and temperature, a cubic spline interpolation algorithm is used to calculate the theoretical on-time and off-time corresponding to the temperature data. The temperature data, theoretical on-time and off-time are combined to obtain the actual on-time and off-time of the IGBT or MOSFET. The actual on-time and off-time are received, and a PWM waveform is generated based on the actual on-time and off-time. The PWM waveform is applied to the gate drive circuit of the power semiconductor device, and the rising and falling edge times of the PWM waveform are adjusted to achieve dynamic adjustment of the switching process of the power semiconductor device. If the power semiconductor device is a bridge arm structure, the dead time is determined based on the actual on-time and off-time to prevent the bridge arm from short-circuiting.

[0026] Specifically, a K-type thermocouple temperature sensor is used to collect real-time temperature data from power semiconductor devices at a sampling frequency of 1kHz. The analog signal is converted to a digital signal using the ADS1115 16-bit ADC and input to a TMS320F28335 DSP-based PCS controller via the SPI interface for processing. Based on a pre-established database of power semiconductor device switching characteristics and temperature mapping, a cubic spline interpolation algorithm is used to calculate the theoretical turn-on and turn-off times at the current temperature. The database stores switching characteristic data at sampling intervals of 10°C over a range of -40°C to 150°C. Preliminary switching timing parameters, including the turn-on time ton and the turn-off time toff, are obtained through interpolation. An adaptive fuzzy control algorithm, combining the real-time temperature measurement data with the preliminary switching timing parameters, dynamically adjusts the actual turn-on and turn-off times of the IGBT or MOSFET. The algorithm uses temperature deviation and temperature change rate as input and outputs the adjustment values ​​for the turn-on and turn-off times. The fuzzy rule base contains 25 rules, defuzzified using the Mamdani inference method and the centroid method. Finally, an optimized PWM drive signal is generated, including the adjusted turn-on time ton_adj and turn-off time toff_adj. The Xilinx Artix-7 series FPGA is used to achieve precise control of the PWM waveform with a clock frequency of 100MHz, achieving a timing accuracy of 1ns. The optimized switching timing is applied to the gate drive circuit of the power semiconductor device, and the dynamic adjustment of the switching process is achieved by adjusting the rising and falling edge times of the PWM waveform. At the same time, the FPGA also implements adaptive adjustment of the dead time to prevent the bridge arm from short-circuiting. The K-type thermocouple temperature sensor is installed on the surface of the IGBT module with a measurement range of -40°C to 800°C and a sensitivity of approximately 41μV / °C. The sampling frequency is set to 1kHz to capture rapid temperature changes. The ADS1115 16-bit ADC converts the temperature signal into a digital quantity with a resolution of

[0027] 0.125°C / LSB. Data is transmitted to the TMS320F28335 DSP via the SPI interface at a rate of 10 Mbps. The DSP pre-stores a switching characteristics database covering -40°C to 150°C, with 10°C intervals, for a total of 20 data points. Using a cubic spline interpolation algorithm, for a measured temperature of 25°C, the turn-on time ton is calculated to be 150 ns, and the turn-off time toff is calculated to be 200 ns. The adaptive fuzzy control algorithm receives as input the temperature deviation (the difference between the target temperature and the measured temperature) and the temperature change rate, which can be a temperature change within 10 ms. The fuzzy rule base contains 25 if-then rules, such as "IF the temperature deviation is positive AND the temperature change rate is positive THEN the turn-on time increases AND the turn-off time increases." After defuzzification using Mamdani reasoning and the centroid method, the obtained values ​​for ton_adj are 155 ns and toff_adj are 210 ns. A Xilinx Artix-7 FPGA (model XC7A100T) receives the adjusted timing parameters from the DSP. Its 100MHz clock provides a basic timing resolution of 10ns. The FPGA achieves 1ns fine-tuning via a digital delay chain, generating precise PWM waveforms. Simultaneously, the FPGA dynamically adjusts the dead time from an initial value of 200ns to 180ns based on switching speed to ensure switching safety. The optimized PWM signal drives the IGBT gate, achieving temperature-adaptive switching control.

[0028] S104. By real-time sampling and analysis of the output voltage and current waveforms of the energy storage converter PCS, the total harmonic distortion rate and power factor index of the output waveform are calculated, and the output waveform quality is evaluated. If the waveform quality index does not meet the preset requirements, the switching timing parameters of the power semiconductor device are further optimized.

[0029] The output voltage and current of the energy storage converter PCS are sampled in real time to obtain sampling data of the voltage and current; based on the sampling data, a fast Fourier transform algorithm is used to perform harmonic analysis to obtain the amplitude and phase of the harmonic components; the total harmonic distortion rate and power factor are calculated based on the amplitude and phase of the harmonic components; it is determined whether the total harmonic distortion rate is greater than a preset total harmonic distortion rate threshold, or whether the power factor is less than a preset power factor threshold; if the total harmonic distortion rate is greater than the preset total harmonic distortion rate threshold, or the power factor is less than the preset power factor threshold, the switching timing parameters of the power semiconductor device are optimized using a particle swarm optimization algorithm; based on the optimized switching timing parameters, a new switching control signal is generated, and the optimized parameters are transmitted to a PWM generator.

[0030] Specifically, an Analog Devices AD7606 high-speed data acquisition card is used to sample the output voltage and current of the energy storage converter PCS in real time. The sampling frequency is set to 100kHz and the sampling accuracy is 16 bits. The sampled data is transmitted to a Texas Instruments TMS320F28379D digital signal processor (TIA) at a rate of 400Mbps via an LVDS high-speed digital interface for subsequent processing. Harmonic analysis of the sampled voltage and current data is performed using a radix-2 Cooley-Tukey fast Fourier transform algorithm. Each analysis uses 8192 data points, and the amplitude and phase of the first 40 harmonic components are calculated. The total harmonic distortion (THD) is calculated using the formula: where Vn is the nth harmonic voltage and V1 is the fundamental voltage. The power factor (PF) is calculated as the ratio of active power (P) to apparent power (S): PF = P / S. The calculated total harmonic distortion (THD) and power factor are compared with pre-set quality requirements. The preset total harmonic distortion thresholds are 3% and 0.95, respectively. If the total harmonic distortion (THD) rate is greater than 3% or the power factor is less than 0.95, the output waveform quality is deemed substandard, triggering the optimization process. The specific parameter values ​​that fail to meet the standards are recorded in the system log for subsequent analysis. Based on the waveform quality assessment results, a particle swarm optimization algorithm is used to optimize the switching timing parameters of the power semiconductor devices. The particle swarm size is set to 50, and the maximum number of iterations is 100. The optimization objective function consists of a weighted sum of the THD rate and the power factor. The optimization variables include the PWM waveform frequency, which ranges from 10kHz to 50kHz, the duty cycle, which ranges from 40% to 60%, and the dead time, which ranges from 0.5μs to 2μs. Through iterative optimization, new switching control signals are generated, and the optimized parameters are transmitted to the PWM generator via the SPI interface. The PCS output is sampled by an AD7606 high-speed data acquisition card at 100kHz, with a voltage range of ±10V and a current range of ±5A, achieving 16-bit resolutions of 0.3mV and 0.15mA, respectively. Sampled data is transmitted to the TMS320F28379D processor via an LVDS interface at 400 Mbps, generating an 8192-point data frame every 10 ms. The processor executes a radix-2 Cooley-Tukey Fast Fourier Transform (FFT) algorithm to calculate 40 harmonic components in the 0-2 kHz range. For example, in a standard 50 Hz system, the detected fundamental voltage is 220 V, the third harmonic is 6.6 V (3%), the fifth harmonic is 4.4 V (2%), and the seventh harmonic is 2.2 V (1%). Based on this, the total harmonic distortion (THD) is calculated to be 3.74%. Simultaneously, the power factor is calculated to be 0.952 based on the active power of 10 kW and the apparent power of 10.5 kVA. These indicators are compared with preset thresholds (THD 3% and power factor 0.95) to trigger the optimization process. The particle swarm algorithm initializes 50 particles, each of which contains three parameters: PWM frequency, duty cycle, and dead time.The optimization objective function, f = 0.6THD + 0.4(1-PF), converged after 78 iterations, yielding the optimal PWM parameters: 20kHz frequency, 52% duty cycle, and 1.2μs dead time. These optimized parameters were transmitted to the PWM generator via an SPI interface at 10Mbps, generating a new switching control signal. This reduced the output waveform THD to 2.8% and increased the power factor to 0.978, meeting quality requirements.

[0031] S105. Use a power analyzer to measure the input power and output power of the energy storage converter PCS, calculate the working efficiency of the energy storage converter PCS, and compare the measured efficiency value with a preset efficiency threshold. If the efficiency is lower than the threshold, reduce the switching loss by adjusting the switching timing parameters of the power semiconductor device.

[0032] The fast Fourier transform method is used to process the voltage and current data at the input and output ends of the energy storage converter PCS to determine the input power and output power; the real-time working efficiency of the energy storage converter PCS is calculated based on the input power and the output power; if the real-time working efficiency is lower than a preset efficiency threshold, the switch timing optimization process is triggered; a genetic algorithm is used to optimize the switch timing parameters of the power semiconductor device, the switch timing parameters including PWM frequency, duty cycle and dead time; a new switch control signal is generated based on the switch timing parameters to reduce switching losses; and the working state parameters of the energy storage converter PCS are obtained, the working state parameters including input voltage, input current, output power, switching frequency and duty cycle.

[0033] Specifically, a Yokogawa WT5000 high-precision power analyzer was used to sample the voltage and current at the input and output terminals of the energy storage converter PCS in real time. The sampling frequency was set to 500kHz, achieving a measurement accuracy of 0.1%. The voltage range was 0-1000V, and the current range was 0-100A. The collected data was transmitted at a rate of 1MB / s via a GP-IB high-speed digital interface to a data processing unit based on an Intel Core i7 processor. The data processing unit processed the sampled data using a fast Fourier transform based on the Cooley-Tukey algorithm, processing 16,384 data points at a time. Calculate the effective values ​​of input and output power: Input power Pin = Vin_rms * Iin_rms * PF_in, and output power Pout = Vout_rms * Iout_rms * PF_out. Vin_rms is the effective value of the input voltage, representing the RMS value of the input voltage; Iin_rms is the effective value of the input current, representing the RMS value of the input current; PFin is the power factor at the input; Vout_rms is the effective value of the output voltage, representing the RMS value of the output voltage; Iout_rms is the effective value of the output current, representing the RMS value of the output current; PFout is the power factor at the output. The real-time operating efficiency of the energy storage converter (PCS) is calculated using the efficiency calculation formula η = Pout / Pin * 100%. The calculated efficiency value is compared with a preset efficiency threshold of 97%. If the efficiency is lower than 97%, the switching timing optimization process is triggered. Simultaneously, current operating state parameters are recorded in the system log, including input voltage (accurate to 0.1V), input current (accurate to 0.01A), output power (accurate to 1W), switching frequency, and duty cycle. Based on these recorded operating state parameters, an improved genetic algorithm is used to optimize the switching timing parameters of power semiconductor devices. The population size is set to 100, with a maximum number of iterations of 500. The chromosome encoding includes the PWM frequency, which ranges from 10kHz to 50kHz in 1kHz steps, the duty cycle, which ranges from 40% to 60% in 0.1% steps, and the dead time, which ranges from 0.5μs to 2μs in 0.1μs steps. The fitness function is set to efficiency η, and a new generation of populations is generated through roulette wheel selection, single-point crossover, and uniform mutation. After iterative optimization, the parameters of the optimal individual are transmitted to the PWM generator via the SPI interface to generate new switching control signals to reduce switching losses. A Yokogawa WT5000 power analyzer samples the PCS, measuring 750V voltage and 80A current at the input and 380V voltage and 150A current at the output. The instrument acquires 16,384 data points at a sampling frequency of 500kHz. The data is transferred to an Intel Core i7 processor via a GP-IB interface at a rate of 1MB / s.The processor executes the Cooley-Tukey FFT algorithm to calculate an input power of 59.4 kW with a power factor of 0.99, and an output power of 57.0 kW with a power factor of 1.0. According to η = 57.0 / 59.4*100%, an efficiency of 96.0% is obtained. Since this efficiency is lower than the preset threshold of 97%, an optimization process is triggered. The current state is recorded, with an input of 750.2 V / 79.98 A, an output of 57023 W, a switching frequency of 20 kHz, and a duty cycle of 50%. The genetic algorithm initializes 100 chromosomes, each containing three genes of PWM frequency, duty cycle, and dead time, such as {25 kHz, 52.5%, 1.2 μs}. After 350 iterations, the optimal individual is {23 kHz, 51.8%, 0.9 μs}, and the predicted efficiency is increased to 97.2%. The optimized parameters are transmitted to the PWM generator TIUCC27714A at a rate of 10 Mbps through the SPI interface to update the IGBT drive signal. After completing one round of optimization, the power analyzer re-measures and verifies that the efficiency is increased to 97.1%, reaching the expected target.

[0034] S106, for different working conditions of the energy storage converter PCS, including light load, heavy load, overload, the corresponding power semiconductor device optimal switching time sequence parameter table is established, the switching time sequence parameter table clearly lists the optimal turn-on time and turn-off time under each load condition, including different turn-on time and turn-off time corresponding to light load, heavy load, overload, during the operation of the energy storage converter PCS, the corresponding switching time sequence parameter is switched in real time according to the load condition.

[0035] The input and output power of the energy storage converter PCS under different load conditions is measured to obtain power measurement data. According to the power measurement data, the switching time sequence parameters of the power semiconductor device in each load interval are optimized using a particle swarm optimization algorithm to obtain optimal switching time sequence parameters. The optimal switching time sequence parameters are arranged into a two-dimensional parameter table and stored in binary format in the memory of the PCS controller to form a parameter lookup table. The output voltage and current are sampled, and the current output power is calculated in real time according to the sampling data. If the calculated current output power changes, the optimal switching time sequence parameters corresponding to the current output power are read from the parameter lookup table. The driving signal is generated according to the read optimal switching time sequence parameters, and the driving signal is applied to the IGBT drive circuit to realize real-time adjustment of the switching time sequence.

[0036] Specifically, a Yokogawa WT5000 high-precision power analyzer was used to measure the input and output power of the energy storage converter PCS under different load conditions. The sampling frequency was set to 500kHz, achieving a measurement accuracy of 0.1%. Based on the measurement results, the load conditions were divided into three ranges: light load (0-30% rated power), heavy load (30%-80% rated power), and overload (80%-120% rated power). Within each range, multiple sampling points were selected for detailed measurements with a step size of 10% rated power. A particle swarm optimization algorithm was used to optimize the switching timing parameters of the power semiconductor devices in each load range. The particle swarm size was set to 50, and the maximum number of iterations was set to 200. The optimization objective function was f = w1*(1-η)+w2*THD, where η is efficiency, THD is total harmonic distortion, and w1 and w2 are weighting coefficients. The optimization variables include the turn-on time, which ranges from 100ns to 500ns, and the turn-off time, which ranges from 200ns to 800ns. Through iterative optimization, the optimal switching timing parameters for each load range are obtained. The optimized switching timing parameters are organized into a two-dimensional parameter table, with the horizontal axis representing load power (0%-120%, in 10% increments) and the vertical axis representing parameter type, including turn-on and turn-off times. The table clearly lists the optimal turn-on and turn-off times for each load point. The parameter table is stored in binary format in the PCS controller's Flash memory, with each parameter occupying 16 bits, for a total storage space of approximately 1KB. During operation of the energy storage converter PCS, the output voltage and current are sampled using an ADS1256 24-bit ADC at a sampling frequency of 10kHz. The FPGA calculates the current output power in real time and maps the power value to the corresponding load range in the parameter table. Based on the mapping results, the corresponding optimal switching timing parameters are read from the Flash memory. The PWM module implemented in the FPGA applies the new switching timing parameters to the IGBT driver circuit, enabling real-time adjustment of the switching timing. The adjustment process is completed within 100μs. A Yokogawa WT5000 power analyzer measured a 1MW energy storage PCS, sampling every 10% over a power range of 0-120%. Efficiency was measured at 92-95% in the light-load range of 0-30%, 96-98% in the heavy-load range of 30-80%, and 94-97% in the overload range of 80-120%. A particle swarm optimization algorithm was used with 50 particles and 200 iterations, using optimization function weights w1 = 0.7 and w2 = 0.3. At 30% load, the optimized turn-on time was 250ns and the turn-off time was 450ns; at 80% load, the turn-on time was 180ns and the turn-off time was 320ns; and at 120% load, the turn-on time was 150ns and the turn-off time was 280ns. The optimized parameter table is stored in the 128KB AT25SF641 flash memory, occupying 1KB of memory.During operation, the ADS1256 samples the output voltage (0-1000V) and current (0-1000A) at a 10kHz frequency. The Xilinx Artix-7 FPGA (XC7A100T) calculates the output power every 100μs. If the output power is detected to be 750kW, the FPGA maps it to the 75% load point in the parameter table and reads the corresponding turn-on time of 190ns and turn-off time of 340ns from flash memory. The FPGA's built-in PWM module runs at a 100MHz clock frequency, achieving a turn-on time of 19 clock cycles and a turn-off time of 34 clock cycles. The updated PWM signal is transmitted via optical fiber to the IGBT gate driver circuit, the CONCEPT2SP0320, to adjust the switching timing.

[0037] S107. Based on the switching characteristics, operating temperature, output waveform quality, and working efficiency of the power device of the energy storage converter PCS, a multi-objective optimization model is constructed. The particle swarm optimization algorithm is used to globally optimize the switching timing parameters to obtain the switching timing control strategy with the best overall performance, which is used to guide the debugging and operation of the PCS.

[0038] A power analyzer is used to measure the input and output power of a PCS to obtain PCS input and output power measurement data; an oscilloscope is used to measure the switching waveform of an IGBT to obtain IGBT switching waveform information; an infrared thermal imager is used to monitor the temperature of a power device to obtain power device temperature data; the power measurement data, switching waveform information and temperature data are synchronously collected through a data acquisition card; a multi-objective optimization model is constructed based on the synchronously collected data, and the objective functions of the multi-objective optimization model include switching loss, temperature, total harmonic distortion rate and efficiency; a multi-objective particle swarm algorithm is used to perform global optimization on the multi-objective optimization model, and the optimization variables of the global optimization include switching frequency, dead time and driving voltage; and the switch timing control strategy obtained by the global optimization is applied to a PCS controller.

[0039] Specifically, Yokogawa WT5000 power analyzer was used to measure the input and output power of PCS with a sampling frequency of 1 MHz. LeCroy HDO6104A oscilloscope was used to measure the switching waveform of IGBT with a bandwidth of 500 MHz. Fluke Ti480PRO infrared thermal imager was used to monitor the temperature of power devices with an accuracy of ±1℃. All data were synchronously collected by National Instruments PXIe-6363 data acquisition card with a sampling rate of 1 MS / s. A multi-objective optimization model was constructed with the objective functions f1 = w1*(Psw / Prated) + w2*(Tmax-Tmin) / Trated and f2 = w3*THD + w4*(1-η), where Psw is the switching loss, Prated is the rated power, Tmax and Tmin are the maximum and minimum operating temperatures, Trated is the rated temperature, THD is the total harmonic distortion, and η is the efficiency. The corresponding weight coefficients w1 to w4 were determined by the analytic hierarchy process. An improved MOPSO (multi-objective particle swarm optimization) algorithm was used for global optimization. The improvements included the use of adaptive inertia weight, the introduction of crowding distance mechanism and ε-dominance ε-dominance concept. The particle swarm size was set to 100 and the maximum number of iterations was set to 500. The optimization variables included switching frequency, dead time and drive voltage, with ranges of 5 kHz-20 kHz, 0.5 μs-2 μs and 15 V-20 V, respectively. The optimized switching timing control strategy was applied to the PCS controller based on TITMS320F28379D. The accurate generation of PWM waveform was realized by FPGA with a timing accuracy of 10 ns. During the operation of PCS, performance data were collected every 100 ms, and short-term performance indicators were calculated using the sliding window method. If the performance deviation was detected to exceed the preset threshold, the online fine-tuning process was triggered, and the gradient descent method was used for local optimization of switching parameters. Yokogawa WT5000 sampled the input and output power of 500 kW PCS at a frequency of 1 MHz, and the peak efficiency of 98.2% was measured. LeCroy HDO6104A recorded the IGBT switching waveform with a rise time of 120 ns and a fall time of 180 ns. Fluke Ti480PRO monitored the IGBT temperature with a maximum point of 134℃. NIPXIe-6363 synchronously collected all data with 1 million sampling points within 1 second. In the multi-objective optimization model, the weight coefficients were obtained by the analytic hierarchy process as w1 = 0.3, w2 = 0.2, w3 = 0.3, and w4 = 0.2. The improved MOPSO algorithm was set with 100 particles and converged after 355 iterations. The optimal solution was a switching frequency of 12.5 kHz, a dead time of 0.8 μs, and a drive voltage of 18.2 V. Under this parameter combination, the switching loss was reduced by 15%, the temperature fluctuation was reduced by 5℃, the THD was reduced to 1.8%, and the efficiency was improved by 0.3%.The TITMS320F28379D controller receives the optimized parameters and generates a PWM waveform via a Xilinx Artix-7 FPGA, achieving a timing accuracy of 9.8ns. During operation, data is collected every 100ms, and performance metrics are calculated using a 200ms sliding window. When a sudden increase in THD to 2.2% is detected, online fine-tuning is triggered. The switching frequency is adjusted using a gradient descent method in 0.1kHz steps. After three iterations, THD returns to 1.9%.

[0040] The description of the above embodiments is only used to help understand the technical solutions and core ideas of this application; ordinary technicians in this field should understand that they can still modify the technical solutions recorded in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A debugging method for a PCS energy storage converter, characterized in that: The method comprises: Obtain the type parameters of the power semiconductor devices in the energy storage converter PCS, including the model, rated voltage, and rated current of the IGBT or MOSFET. Based on the parameters of the power semiconductor devices, determine the turn-on time and turn-off time characteristic curves of the power semiconductor devices at different temperatures, and obtain the mapping relationship between the switching characteristics of the power semiconductor devices and temperature; A temperature sensor is used to detect the operating temperature of the power semiconductor device in the energy storage converter PCS in real time, and the detected temperature value is compared with a preset temperature threshold. If the detected temperature exceeds the threshold, it is determined that the power semiconductor device is operating in a high temperature state and the switching timing needs to be adjusted to adapt to the temperature change. The temperature sensor is installed on the surface of the power semiconductor device; Based on the mapping relationship between the switching characteristics and temperature of power semiconductor devices, combined with the real-time temperature detection value, the on-time and off-time of IGBT or MOSFET are dynamically adjusted to optimize the switching timing of power semiconductor devices to adapt to the working conditions at different temperatures. By sampling and analyzing the output voltage and current waveforms of the energy storage converter PCS in real time, the total harmonic distortion rate and power factor of the output waveform are calculated, and the output waveform quality is evaluated. If the waveform quality indicators do not meet the preset requirements, the switching timing parameters of the power semiconductor devices are further optimized; A power analyzer is used to measure the input power and output power of the energy storage converter PCS, calculate the operating efficiency of the energy storage converter PCS, and compare the measured efficiency value with a preset efficiency threshold. If the efficiency is lower than the threshold, the switching timing parameters of the power semiconductor device are adjusted to reduce switching losses. For the different operating conditions of the energy storage converter PCS, including light load, heavy load, and overload, a corresponding optimal switching timing parameter table for power semiconductor devices is established. The switching timing parameter table clearly lists the optimal turn-on and turn-off times under various load conditions, including different turn-on and turn-off times corresponding to light load, heavy load, and overload conditions. During the operation of the energy storage converter PCS, the corresponding switching timing parameters are switched in real time according to the load conditions; Based on the switching characteristics, operating temperature, output waveform quality and working efficiency of the power device of the energy storage converter PCS, a multi-objective optimization model is constructed. The particle swarm optimization algorithm is used to globally optimize the switching timing parameters. The switching timing control strategy with the best comprehensive performance is obtained to guide the debugging and operation of the PCS.

2. The method according to claim 1, characterized in that The method of obtaining the type parameters of the power semiconductor device in the energy storage converter PCS, including the model, rated voltage, and rated current of the IGBT or MOSFET, and determining the turn-on time and turn-off time characteristic curves of the power semiconductor device at different temperatures based on the parameters of the power semiconductor device, and obtaining the switching characteristic and temperature mapping relationship of the power semiconductor device, includes: Obtaining the model, rated voltage, and rated current parameters of the power semiconductor devices in the energy storage converter PCS, and establishing a static parameter database of the power semiconductor devices based on the parameters; Perform temperature sweep tests on power semiconductor devices to obtain a mapping model between switching characteristics and temperature; Calculating the switching loss and conduction loss of the PCS under different working states according to the static parameter database and the mapping relationship model; Importing the switching loss and conduction loss data into thermal simulation software to analyze the junction temperature distribution of the power semiconductor device; If the junction temperature distribution exceeds the maximum junction temperature specified for the device, a particle swarm optimization algorithm is used to optimize the topology of the PCS; A PCS control model is constructed based on the optimized topology structure; and an iterative calculation is performed through the PCS control model to obtain the optimal operating parameters of the PCS.

3. The method according to claim 1, characterized in that The method uses a temperature sensor to detect the operating temperature of the power semiconductor device in the energy storage converter PCS in real time, compares the detected temperature value with a preset temperature threshold, and if the detected temperature exceeds the threshold, it is determined that the power semiconductor device is operating in a high temperature state and the switching timing needs to be adjusted to adapt to the temperature change, including: A thermocouple temperature sensor is installed on the surface of the power semiconductor device. The temperature data is collected in real time at a preset sampling frequency through a data acquisition card. The collected temperature values ​​are converted into digital signals and input into the PCS controller. According to the multi-level temperature thresholds preset in the PCS controller, the detected temperature is compared with the preset temperature threshold in real time by a comparator circuit. If the detected temperature exceeds the first-level warning threshold, the temperature compensation in the PCS controller is triggered; The temperature compensation searches for the corresponding switching frequency in a preset temperature and frequency correspondence table according to the current temperature, and adjusts the duty cycle of the PWM wave according to the temperature and duty cycle correspondence table; If the detected temperature continues to rise and exceeds the secondary warning threshold, the PCS controller further reduces the switching frequency and current stress of the power semiconductor device; If the temperature continues to rise, the preset multi-stage cooling system, including air cooling and water cooling systems, will be activated step by step until the temperature drops or reaches the over-temperature protection threshold.

4. The method according to claim 1, wherein The method dynamically adjusts the on-time and off-time of the IGBT or MOSFET based on the mapping relationship between the switching characteristics and temperature of the power semiconductor device and the acquired real-time temperature detection value, and optimizes the switching timing of the power semiconductor device to adapt to the working conditions at different temperatures, including: Use K-type thermocouple temperature sensor to collect temperature data of power semiconductor devices in real time; Based on the pre-established database of mapping relationship between switching characteristics and temperature of power semiconductor devices, the theoretical turn-on time and turn-off time corresponding to the temperature data are calculated using the cubic spline interpolation algorithm; Combine temperature data, theoretical turn-on time and turn-off time to obtain the actual turn-on time and turn-off time of IGBT or MOSFET; Receive the actual on-time and off-time, and generate a PWM waveform according to the actual on-time and off-time; Applying the PWM waveform to a gate drive circuit of a power semiconductor device, and dynamically adjusting the switching process of the power semiconductor device by adjusting the rising and falling edge times of the PWM waveform; If the power semiconductor device is a bridge arm structure, the dead time is determined according to the actual turn-on time and turn-off time to prevent the bridge arm from short-circuiting.

5. The method according to claim 1, wherein The method comprises: performing real-time sampling and analysis on the output voltage and current waveforms of the energy storage converter PCS, calculating the total harmonic distortion rate and power factor index of the output waveform, evaluating the output waveform quality, and further optimizing the switching timing parameters of the power semiconductor device if the waveform quality index does not meet the preset requirements. The output voltage and current of the energy storage converter PCS are sampled in real time to obtain the sampling data of voltage and current; Performing harmonic analysis on the sampled data using a fast Fourier transform algorithm to obtain the amplitude and phase of the harmonic components; Calculating the total harmonic distortion and power factor based on the amplitude and phase of the harmonic components; Determining whether the total harmonic distortion rate is greater than a preset total harmonic distortion rate threshold, or whether the power factor is less than a preset power factor threshold; If the total harmonic distortion rate is greater than a preset total harmonic distortion rate threshold, or the power factor is less than a preset power factor threshold, optimizing the switching timing parameters of the power semiconductor device using a particle swarm optimization algorithm; According to the optimized switch timing parameters, a new switch control signal is generated, and the optimized parameters are transmitted to the PWM generator.

6. The method according to claim 1, characterized in that The power analyzer is used to measure the input power and output power of the energy storage converter PCS, calculate the working efficiency of the energy storage converter PCS, compare the measured efficiency value with a preset efficiency threshold, and if the efficiency is lower than the threshold, reduce the switching loss by adjusting the switching timing parameters of the power semiconductor device, including: The fast Fourier transform method is used to process the voltage and current data at the input and output ends of the energy storage converter PCS to determine the input power and output power; Calculating the real-time working efficiency of the energy storage converter PCS according to the input power and the output power; If the real-time working efficiency is lower than a preset efficiency threshold, a switch timing optimization process is triggered; Using a genetic algorithm to optimize the switching timing parameters of the power semiconductor device, the switching timing parameters include PWM frequency, duty cycle and dead time; generating a new switch control signal according to the switch timing parameters to reduce switching losses; The operating state parameters of the energy storage converter PCS are obtained, where the operating state parameters include input voltage, input current, output power, switching frequency, and duty cycle.

7. The method according to claim 1, characterized in that For different operating conditions of the energy storage converter PCS, including light load, heavy load, and overload, a corresponding optimal switching timing parameter table of power semiconductor devices is established. The switching timing parameter table clearly lists the optimal turn-on time and turn-off time under each load condition, including different turn-on times and turn-off times corresponding to light load, heavy load, and overload conditions. During the operation of the energy storage converter PCS, the corresponding switching timing parameters are switched in real time according to the load conditions, including: Measure the input and output power of the energy storage converter PCS under different load conditions to obtain power measurement data; Optimizing the switching timing parameters of the power semiconductor device in each load interval using a particle swarm optimization algorithm based on the power measurement data to obtain optimal switching timing parameters; Arranging the optimal switch timing parameters into a two-dimensional parameter table and storing the table in a memory of the PCS controller in a binary format to form a parameter lookup table; By sampling the output voltage and current, the current output power is calculated in real time based on the sampled data; If the calculated current output power changes, the optimal switch timing parameters corresponding to the current output power are read from the parameter lookup table; A drive signal is generated according to the read optimal switch timing parameters, and the drive signal is applied to the IGBT drive circuit to achieve real-time adjustment of the switch timing.

8. The method according to claim 1, characterized in that The multi-objective optimization model is constructed based on the power device switching characteristics, operating temperature, output waveform quality and working efficiency of the energy storage converter PCS. The particle swarm optimization algorithm is used to perform global optimization on the switch timing parameters to obtain the switch timing control strategy with the best overall performance, which is used to guide the debugging and operation of the PCS. The model includes: Use a power analyzer to measure the input and output power of the PCS to obtain PCS input and output power measurement data; Use an oscilloscope to measure the IGBT switching waveform and obtain the IGBT switching waveform information; Use infrared thermal imager to monitor the temperature of power devices and obtain power device temperature data; Synchronously collecting the power measurement data, switching waveform information and temperature data through a data acquisition card; Constructing a multi-objective optimization model based on the synchronously collected data, wherein the objective functions of the multi-objective optimization model include switching loss, temperature, total harmonic distortion, and efficiency; A multi-objective particle swarm algorithm is used to perform global optimization on the multi-objective optimization model, wherein the optimization variables of the global optimization include switching frequency, dead time and driving voltage; The switch timing control strategy obtained by global optimization is applied to the PCS controller.

Images