Power supply quality analysis model based on neural network algorithm

Abstract

The invention provides a power supply quality analysis model based on a neural network algorithm, and relates to the field of power supply quality analysis. A power management module can provide temporary power support for a whole analysis system when commercial power is interrupted; the data acquisition module is used for converting the current signal into a digital signal and acquiring the digital signal; the data preprocessing module is used for amplifying and filtering the collected signals; the feature extraction module is used for calculating time domain features and frequency domain features of the received data; and the analysis and evaluation module calculates the preprocessed data and generates a graph, thereby solving the problems that when an existing power supply quality analysis model based on a neural network algorithm is used for analyzing the power supply quality, if the collected power supply quality data is interfered, the power supply quality analysis model is very easy to generate errors in the analysis process, and the analysis accuracy is poor. Therefore, the problem that the analysis quality is affected badly is solved, and the accuracy of an analysis result is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of power supply quality analysis, and in particular to a power supply quality analysis model based on a neural network algorithm. Background Art

[0002] As a computational model that simulates the neural structure of the human brain, neural network algorithms have demonstrated excellent capabilities in solving various complex problems. In modern society, power supply is undoubtedly the key lifeline for the normal operation of various fields. The quality of power supply is directly related to the quality and efficiency of industrial production, commercial operations and residents' lives. With the continuous advancement of science and technology and the rapid development of social economy, the complexity of the power system is increasing day by day, and the power supply quality is facing unprecedented challenges. In this context, the requirements for accuracy and timeliness when analyzing and evaluating the power supply quality are also rising.

[0003] According to the application number: CN201010102923.X, a method for analyzing the available power supply capacity of a medium-voltage distribution network is disclosed, which includes: establishing an available power supply capacity model; analyzing the communication relationship between the main transformers in the research area to form a main transformer communication relationship matrix Llink; dividing the Llink matrix into blocks according to the difference between intra-station communication and inter-station communication, and correcting the main transformer communication relationship matrix Llink; defining the capacity matrix R, each element of which represents the capacity of the main transformer corresponding to the communication center, and correcting the R matrix; calculating the available power supply capacity of the system; locating weak points; and obtaining critical capacity. The method for analyzing the available power supply capacity of a medium-voltage distribution network of the present invention proposes a calculation model and a calculation and analysis method for the available power supply capacity, and on the basis of the available power supply capacity, proposes an analysis method for quickly locating the weak points in the grid that restrict the available power supply capacity through the transfer channel bottleneck matrix, and obtaining the minimum component capacity required for the connection line through the critical capacity matrix.

[0004] However, when the existing power supply quality analysis model based on the neural network algorithm is used to analyze power supply quality, if the collected power supply quality data is disturbed, it is very easy to cause errors in the power supply quality analysis model during the analysis process, thereby adversely affecting the analysis quality. Summary of the invention

[0005] In view of this, the present invention provides a power supply quality analysis model based on a neural network algorithm, which has the advantages of ensuring the continuity of power supply quality analysis work, ensuring the accuracy of analysis results, and ensuring the security of sensitive information such as power supply quality data and analysis results.

[0006] The present invention provides a power supply quality analysis model based on a neural network algorithm, which specifically includes a power management module, a data acquisition module, a data preprocessing module, a feature extraction module, a data encryption module, an analysis and evaluation module and a visualization module; the power management module can provide temporary power support for the entire analysis system when the city power is interrupted; the data acquisition module converts the current signal into a digital signal and collects it; the data preprocessing module amplifies and filters the collected signal; the feature extraction module calculates the time domain characteristics and frequency domain characteristics of the received data; the data encryption module encrypts the collected power supply quality data, analysis results and other sensitive information; the analysis and evaluation module calculates the preprocessed data and generates graphics; the visualization module presents the generated graphics.

[0007] Furthermore, the power management module includes: a UPS component and a power filter; the UPS component is composed of a battery pack, a charging circuit, and an inverter. The battery pack stores electrical energy, the charging circuit ensures that the battery is in a good charging state, and the inverter converts the DC power of the battery into AC power for the system; the power filter is composed of a common-mode inductor, a differential-mode inductor, a capacitor, and a resistor; the common-mode inductor can suppress high-frequency common-mode interference, and the differential-mode inductor suppresses differential-mode interference.

[0008] Furthermore, the data acquisition module includes: a sensor component and a data acquisition card; the sensor component is composed of a voltage sensor, a current sensor, and a power quality sensor, the voltage sensor converts a high voltage into a low voltage signal that can be measured, the current sensor converts a large current into a small current signal in proportion for measurement, and the power quality sensor determines the amplitude and phase of each harmonic; the data acquisition card is composed of an analog-to-digital conversion component, a data cache component, and a communication interface component, the analog-to-digital conversion component converts the analog signal collected by the sensor component into a digital signal, and the data cache component enables the data acquisition card to have data cache space.

[0009] Furthermore, the data preprocessing module includes: a signal conditioning circuit device and a programmable logic device; the signal conditioning circuit device is composed of an amplifying circuit component, a filtering circuit component, and an isolation circuit component, the amplifying circuit component amplifies the signal output by the sensor component, the filtering circuit component filters the signal collected by the sensor component, and the isolation circuit component uses an optical signal to transmit an electrical signal.

[0010] Furthermore, the feature extraction module includes: a digital signal processor; the digital signal processor is electrically connected to the signal conditioning circuit device.

[0011] Furthermore, the data encryption module includes: a hardware encryption chip and a security authentication component; the hardware encryption chip encrypts the collected power supply quality data and sensitive information of the analysis results based on the AES algorithm.

[0012] Furthermore, the analysis and evaluation module includes: a central processing unit, a graphics processing unit and a storage device; the storage device is composed of a random access memory and a hard disk, the random access memory ensures fast reading and writing of data, and the hard disk stores historical data, analysis results and related model parameters of power supply quality analysis.

[0013] Furthermore, the visualization module includes: a graphics display card and a display terminal; the graphics display card is electrically connected to a graphics processing unit; and the display terminal is electrically connected to the graphics display card.

[0014] Beneficial Effects The present invention realizes the detection and analysis of power supply quality through a power management module, a data acquisition module, a data preprocessing module, a feature extraction module, an analysis and evaluation module, a data encryption module and a visualization module.

[0015] In addition, through the setting of the power management module, the stable operation of the power supply quality analysis system is achieved, avoiding data loss or equipment damage when the power grid fails and power outages, and ensuring the continuity of the power supply quality analysis work.

[0016] In addition, through the setting of the data preprocessing module, stable processing and calculation of the collected data is achieved, avoiding external electromagnetic interference and ensuring the accuracy of the analysis results.

[0017] In addition, through the setting of the data encryption module, the collected power supply quality data, analysis results and other sensitive information can be encrypted, ensuring the security of sensitive information such as power supply quality data and analysis results. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the drawings of the embodiment are briefly introduced below.

[0019] The drawings described below are only related to some embodiments of the present invention, but are not intended to limit the present invention.

[0020] In the attached picture: Figure 1 It is a structural schematic diagram of a power supply quality analysis model based on a neural network algorithm of the present invention.

[0021] Figure 2 It is a structural diagram of data collection in the first embodiment of the present invention.

[0022] Figure 3It is a structural diagram of data processing in the first embodiment of the present invention.

[0023] Figure 4 Schematic diagram of the structure of the visualization module of the first embodiment of the present invention Figure 5 It is an operation flow chart of the power management module of the first embodiment of the present invention.

[0024] Figure 6 It is an operation flow chart of the data acquisition module of the first embodiment of the present invention.

[0025] Figure 7 It is an operation flow chart of the data preprocessing module of the first embodiment of the present invention.

[0026] Figure 8 It is an operation flow chart of the digital signal processor and the data encryption module of the first embodiment of the present invention.

[0027] Fig. 9 It is an operation flow chart of the analysis and evaluation module and the visualization module of the first embodiment of the present invention.

[0028] Reference numerals list 1. Power management module; 101. UPS components; 1011. Battery pack; 1012. Charging circuit; 1013. Inverter; 102. Power filter; 1021. Common mode inductor; 1022. Differential mode inductor; 1023. Capacitor; 1024. Resistor; 2. Data acquisition module; 201. Sensor components; 2011. Voltage sensor; 2012. Current sensor; 2013. Power quality sensor; 202. Data acquisition card; 2021. Analog-to-digital conversion component; 2022. Data cache component; 2023. Communication Interface component; 3. Data preprocessing module; 301. Signal conditioning circuit device; 3011. Amplifying circuit component; 3012. Filtering circuit component; 3013. Isolation circuit component; 302. Programmable logic device; 4. Feature extraction module; 401. Digital signal processor; 5. Data encryption module; 501. Hardware encryption chip; 502. Security authentication component; 6. Analysis and evaluation module; 601. Central processing unit; 602. Graphics processing unit; 603. Storage device; 6031. Random access memory; 6032. Hard disk; 7. Visualization module; 701. Graphics display card; 702. Display terminal. DETAILED DESCRIPTION

[0029] Embodiment 1: Please refer to Figure 1-Figure 9 As shown: The present invention provides a power supply quality analysis model based on a neural network algorithm, comprising a power management module 1, a data acquisition module 2, a data preprocessing module 3, a feature extraction module 4, a data encryption module 5, an analysis and evaluation module 6 and a visualization module 7; the power management module 1 can provide temporary power support for the entire analysis system when the city power is interrupted; the data acquisition module 2 converts the current signal into a digital signal and collects it; the data preprocessing module 3 amplifies and filters the collected signal; the feature extraction module 4 calculates the time domain characteristics and frequency domain characteristics of the received data; the data encryption module 5 encrypts the collected power supply quality data, analysis results and other sensitive information; the analysis and evaluation module 6 calculates the preprocessed data and generates graphics; the visualization module 7 presents the generated graphics.

[0030] The power management module 1 includes: a UPS component 101 and a power filter 102; the UPS component 101 is composed of a battery pack 1011, a charging circuit 1012, and an inverter 1013. The battery pack 1011 stores electrical energy, the charging circuit 1012 ensures that the battery is in a good charging state, and the inverter 1013 converts the direct current of the battery into alternating current for use by the system. When a power grid fails and a power outage occurs, the UPS component 101 can continue to supply power to the power quality analysis equipment to avoid data loss and equipment damage, and ensure the continuity of the power quality analysis work; the power filter 102 is composed of a common-mode inductor 1021, a differential-mode inductor 1022, a capacitor 1023, and a resistor 1024. The common-mode inductor 1021 can suppress high-frequency common-mode interference, and the differential-mode inductor 1022 is connected in series in the live wire or the neutral wire to increase the inductive reactance to the differential-mode current, thereby suppressing the differential-mode interference. The capacitor 1023 can prevent the circuit from resonating in the filter circuit and improve the stability of the filtering effect.

[0031] The data acquisition module 2 includes: a sensor component 201 and a data acquisition card 202; the sensor component 201 is composed of a voltage sensor 2011, a current sensor 2012, and a power quality sensor 2013. The voltage sensor 2011 converts a high voltage into a measurable low voltage signal based on the principle of electromagnetic induction. The current sensor 2012 converts a large current into a small current signal in proportion for measurement through electromagnetic induction. The power quality sensor 2013 determines the amplitude and phase of each harmonic by performing spectrum analysis on the voltage or current signal; the data acquisition card 202 is composed of an analog-to-digital conversion component 2021, a data cache component 2022, and a communication interface component 2023. The analog-to-digital conversion component 2021 converts the analog signal collected by the sensor component 201 into a digital signal. During the collection process, the data cache component 2022 enables the data acquisition card 202 to have a certain data cache space. When the collected data cannot be transmitted to the host computer in time, it can be stored in the cache first to avoid data loss. The communication interface component 2023 transmits the collected data to the host computer through a USB interface, an Ethernet interface, and an RS-485 interface. 485 interface and other communication interfaces transmit the collected data to the host computer.

[0032] Among them, the data preprocessing module 3 includes: a signal conditioning circuit device 301 and a programmable logic device 302; the signal conditioning circuit device 301 is composed of an amplifier circuit component 3011, a filter circuit component 3012, and an isolation circuit component 3013. The amplifier circuit component 3011 amplifies the signal output by the sensor component 201 to a suitable amplitude so that the subsequent acquisition equipment can accurately identify and process it. The filter circuit component 3012 filters the signal collected by the sensor component 201 to remove noise and interference signals. The isolation circuit component 3013 uses optical signals to transmit electrical signals to electrically isolate the input and output sides to ensure the safety and reliability of the data acquisition system; the programmable logic device 302 implements threshold-based abnormal value detection through programming. A large number of programmable logic units inside the FPGA can process data in parallel to quickly detect voltage and current data values ​​that exceed the normal range. The programmable logic device 302 can quickly start and execute neural network calculations to provide real-time power supply quality analysis results.

[0033] Among them, the feature extraction module 4 includes: a digital signal processor 401; the digital signal processor 401 has a dedicated hardware multiplier and fast computing capabilities, which can efficiently execute digital signal processing algorithms. In feature extraction, it can be used to quickly calculate time domain features and frequency domain features, and can process a large amount of voltage and current sampling data in a short time, complete complex mathematical operations, and meet real-time feature extraction requirements.

[0034] Among them, the data encryption module 5 includes: a hardware encryption chip 501 and a security authentication component 502; the hardware encryption chip 501 encrypts the collected power supply quality data, analysis results and other sensitive information, the encryption algorithm encrypts the data based on the AES algorithm, and the security authentication component 502 authenticates the user or device accessing the power supply quality analysis system.

[0035] The analysis and evaluation module 6 includes: a central processing unit 601, a graphics processing unit 602 and a storage device 603; when the central processing unit 601 calculates the power supply quality index, the CPU can perform calculations according to the calculation formulas specified in the relevant standards based on the collected and pre-processed data. The voltage qualification rate calculation formula is: voltage qualification rate = voltage qualification points / total monitoring points × 100%, and the frequency qualification rate calculation formula is: frequency qualification rate = frequency qualification time / total monitoring time × 100%, the neural network analyzes the input power supply quality data under the coordination of the central processor 601; when the graphics processing unit 602 is training the neural network model for predicting the voltage fluctuation trend, the graphics processing unit 602 can calculate the weight update of the neural network in parallel to improve the model training efficiency; the storage device 603 is composed of a random access memory 6031 and a hard disk 6032. When calculating the power supply quality index, the random access memory 6031 needs to quickly access and process the collected voltage, current and other data. The random access memory 6031 can ensure fast reading and writing of data. The hard disk 6032 stores the historical data, analysis results and related model parameters of the power supply quality analysis. When performing trend analysis, a large amount of historical data is required as a basis, and these data can be stored in the hard disk 6032.

[0036] Among them, the visualization module 7 includes: a graphics display card 701 and a display terminal 702; the graphics display card 701 displays the analysis results in the form of intuitive graphics, charts, etc., the graphics display card 701 can quickly generate high-quality visualization images, and the display terminal 702 can display the visualization images generated by the graphics display card 701.

[0037] The specific usage and function of this embodiment are as follows: the battery pack 1011 stores electrical energy, the charging circuit 1012 ensures that the battery is in a good charging state, the inverter 1013 converts the direct current of the battery into alternating current for use by the system, and when a power outage occurs due to a power failure in the power grid, the UPS component 101 can continue to supply power to the power quality analysis equipment to avoid data loss and equipment damage, and ensure the continuity of the power quality analysis work. The power filter 102 can suppress common-mode interference and differential-mode interference, and prevent the circuit from resonating, thereby improving the stability of the filtering effect. The voltage sensor 2011 converts high voltage into AC based on the principle of electromagnetic induction. The current sensor 2012 converts the large current into a small current signal in proportion for measurement through electromagnetic induction. The power quality sensor 2013 determines the amplitude and phase of each harmonic by performing spectrum analysis on the voltage or current signal. The analog-to-digital conversion component 2021 converts the analog signal collected by the sensor component 201 into a digital signal. During the collection process, the data cache component 2022 enables the data acquisition card 202 to have a certain data cache space. When the collected data cannot be transmitted to the host computer in time, it can be stored in the cache first to avoid data loss. The communication interface component 2023 transmits the data to the host computer through the USB. The communication interfaces such as the interface, Ethernet interface, and RS-485 interface transmit the collected data to the host computer. The amplifier circuit component 3011 amplifies the signal output by the sensor component 201 to a suitable amplitude so that the subsequent acquisition equipment can accurately identify and process it. The filter circuit component 3012 filters the signal collected by the sensor component 201 to remove noise and interference signals. The isolation circuit component 3013 uses optical signals to transmit electrical signals to electrically isolate the input and output sides to ensure the safety and reliability of the data acquisition system. The programmable logic device 302 realizes threshold-based abnormal value detection through programming. The FPGA A large number of internal programmable logic units can process data in parallel and quickly detect voltage and current data values ​​that exceed the normal range. The digital signal processor 401 has a dedicated hardware multiplier and fast computing capability, and can efficiently execute digital signal processing algorithms. In feature extraction, it can be used to quickly calculate time domain features and frequency domain features, and can process a large amount of voltage and current sampling data in a short time, complete complex mathematical operations, and meet the needs of real-time feature extraction. The hardware encryption chip 501 encrypts the collected power quality data, analysis results and other sensitive information. The security authentication component 502 authenticates the user or device accessing the power quality analysis system. When the central processor 601 calculates the power quality index, the CPU can calculate according to the calculation formula specified in the relevant standards based on the collected and pre-processed data. The voltage qualification rate calculation formula is: voltage qualification rate = voltage qualification points / total number of monitoring points × 100%, and the frequency qualification rate calculation formula is: frequency qualification rate = frequency qualification time / total monitoring time × 100%;When the graphics processing unit 602 trains the neural network model for predicting the voltage fluctuation trend, the graphics processing unit 602 can calculate the weight update of the neural network in parallel to improve the model training efficiency. When calculating the power supply quality index, the random access memory 6031 needs to quickly access and process the collected voltage, current and other data. The random access memory 6031 can ensure the fast reading and writing of data. The hard disk 6032 stores the historical data, analysis results and related model parameters of the power supply quality analysis. When performing trend analysis, a large amount of historical data is required as a basis. These data can be stored in the hard disk 6032. The graphics display card 701 displays the analysis results in the form of intuitive graphics, charts, etc. The graphics display card 701 can quickly generate high-quality visual images, and the display terminal 702 can display the visual images generated by the graphics display card 701. ;

[0038] In this article, there are a few points to note: 1. The drawings of the embodiments of the present disclosure only involve structures related to the embodiments of the present disclosure, and other structures may refer to general designs.

[0039] 2. In the absence of conflict, the embodiments of the present disclosure and the features therein may be combined with each other to obtain new embodiments.

[0040] The above are only specific embodiments of the present disclosure, but the protection scope of the present disclosure is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present disclosure, which should be included in the protection scope of the present disclosure. Therefore, the protection scope of the present disclosure should be based on the protection scope of the claims.

Claims

1. A power quality analysis model based on a neural network algorithm, characterized in that: The system comprises a power management module (1), a data acquisition module (2), a data preprocessing module (3), a feature extraction module (4), a data encryption module (5), an analysis and evaluation module (6) and a visualization module (7); the power management module (1) can provide temporary power support for the entire analysis system when the city power is interrupted; the data acquisition module (2) converts the current signal into a digital signal and acquires the signal; the data preprocessing module (3) amplifies and filters the acquired signal; the feature extraction module (4) calculates the time domain features and frequency domain features of the received data; the data encryption module (5) encrypts the acquired power quality data, analysis results and other sensitive information; the analysis and evaluation module (6) calculates the preprocessed data and generates graphics; and the visualization module (7) presents the generated graphics.

2. A power quality analysis model based on a neural network algorithm as claimed in claim 1, characterized in that: The power management module (1) comprises: a UPS component (101) and a power filter (102); the UPS component (101) is composed of a battery pack (1011), a charging circuit (1012), and an inverter (1013); the battery pack (1011) stores electric energy; the charging circuit (1012) ensures that the battery is in a good charging state; and the inverter (1013) converts the direct current of the battery into alternating current for use by the system; the power filter (102) is composed of a common-mode inductor (1021), a differential-mode inductor (1022), a capacitor (1023), and a resistor (1024); the common-mode inductor (1021) can suppress high-frequency common-mode interference, and the differential-mode inductor (1022) suppresses differential-mode interference.

3. A power quality analysis model based on a neural network algorithm as claimed in claim 1, characterized in that: The data acquisition module (2) comprises: a sensor component (201) and a data acquisition card (202); the sensor component (201) is composed of a voltage sensor (2011), a current sensor (2012), and a power quality sensor (2013); the voltage sensor (2011) converts a high voltage into a low voltage signal that can be measured; the current sensor (2012) converts a large current into a small current signal in proportion for measurement; and the power quality sensor (2013) determines the amplitude and phase of each harmonic; the data acquisition card (202) is composed of an analog-to-digital conversion component (2021), a data cache component (2022), and a communication interface component (2023); the analog-to-digital conversion component (2021) converts an analog signal collected by the sensor component (201) into a digital signal; and the data cache component (2022) enables the data acquisition card (202) to have data cache space.

4. A power quality analysis model based on a neural network algorithm as claimed in claim 3, characterized in that: The data preprocessing module (3) comprises: a signal conditioning circuit device (301) and a programmable logic device (302); the signal conditioning circuit device (301) is composed of an amplifying circuit component (3011), a filtering circuit component (3012), and an isolating circuit component (3013); the amplifying circuit component (3011) amplifies the signal output by the sensor component (201); the filtering circuit component (3012) performs filtering processing on the signal collected by the sensor component (201); and the isolating circuit component (3013) transmits the electrical signal using an optical signal.

5. A power quality analysis model based on a neural network algorithm as claimed in claim 4, characterized in that: The feature extraction module (4) comprises: a digital signal processor (401); the digital signal processor (401) is electrically connected to the signal conditioning circuit device (301).

6. A power quality analysis model based on a neural network algorithm as claimed in claim 1, characterized in that: The data encryption module (5) comprises: a hardware encryption chip (501) and a security authentication component (502); the hardware encryption chip (501) encrypts the collected power supply quality data and analysis result sensitive information based on the AES algorithm.

7. The power quality analysis model based on a neural network algorithm as claimed in claim 1, characterized in that: The analysis and evaluation module (6) comprises: a central processing unit (601), a graphics processing unit (602) and a storage device (603); the storage device (603) is composed of a random access memory (6031) and a hard disk (6032); the random access memory (6031) ensures fast reading and writing of data, and the hard disk (6032) stores historical data of power supply quality analysis, analysis results and related model parameters.

8. A power quality analysis model based on a neural network algorithm as claimed in claim 6, characterized in that: The visualization module (7) comprises: a graphics display card (701) and a display terminal (702); the graphics display card (701) is electrically connected to a graphics processing unit (602); and the display terminal (702) is electrically connected to the graphics display card (701).

Images