An energy consumption collection, detection and analysis cloud platform based on the Internet of Things

Abstract

The present application relates to the field of electrical signal measurement technology, and in particular to an energy consumption collection, detection and analysis cloud platform based on the Internet of Things. It includes a sensing unit, an edge computing unit, an identification unit, a sending unit and a cloud server. The sensing unit measures the energy consumption data of the equipment in real time and draws a time series curve graph. Based on the curve graph, multiple target data are sampled and divided into data groups. The edge computing unit identifies whether there is noise in three adjacent target data, and performs correction when there is noise to obtain correction data. The identification unit performs frequency domain transformation based on the target data and the correction data to obtain the frequency domain transformation result, and analyzes the equipment fault status accordingly. The sending unit compresses the frequency domain transformation result when the equipment fault status is normal and sends it to the cloud server. The cloud server restores the energy consumption data and analyzes and obtains the energy consumption analysis result. The present invention solves the problems of low energy consumption data collection accuracy and low processing efficiency, and improves the energy consumption data collection accuracy and processing efficiency.

Description

Technical Field

[0001] The present application relates to the field of electrical signal measurement technology, and in particular to an energy consumption collection, detection and analysis cloud platform based on the Internet of Things. Background Art

[0002] In today's society, with the advancement of science and technology and rising awareness of energy conservation and environmental protection, accurate monitoring and analysis of energy consumption have become increasingly important. Existing energy consumption monitoring technologies often use traditional data collection methods, which suffer from issues such as incomplete data collection, low transmission efficiency, and limited analysis capabilities. Similar prior art includes Chinese patent application number CN119891559A, which proposes an intelligent load collection system based on a mixed-use scenario for a station power system. The system includes a load fluctuation analysis module, a reception frequency adjustment module, an anomaly threshold adjustment module, and an abnormal load identification module. This system deploys sensors from differentiated devices to monitor power loads in real time. Using a cloud server to receive and process power load data in real time, it eliminates data noise caused by the start and stop of individual devices, enabling accurate assessment of power load fluctuations. Dynamic adjustment of the data reception frequency makes data collection more efficient, reduces the transmission of invalid data, and improves the speed and accuracy of data processing. The system analyzes the average power load values ​​of devices based on historical data and adjusts the anomaly detection threshold accordingly, improving the ability and accuracy to identify power load anomalies. This helps prevent power system failures and assists power operators in subsequent energy management optimization. In addition, similar prior art includes a Chinese patent with publication number CN113884758A, which provides a direct current electric energy metering method, device, equipment and storage medium. The method includes obtaining original sampling data of electric energy signals; performing wavelet transform on the original sampling data according to a preset wavelet basis and number of decomposition layers to obtain wavelet coefficients corresponding to each decomposition layer; setting a threshold value and a threshold function for the wavelet coefficients corresponding to each decomposition layer, quantizing the wavelet coefficients corresponding to each decomposition layer, and obtaining quantized wavelet coefficients; reconstructing the quantized wavelet coefficients to obtain a denoised signal; performing ripple error correction on the denoised signal to obtain corrected data; and integrating the corrected data using a composite trapezoidal integration method to obtain target electric energy data, thereby overcoming the disadvantage that the window size changes with frequency, filtering is performed within the entire frequency range, and improving the accuracy of output metering. Both of the above patent documents solve the problem of collecting electric energy signals, but there are problems of low signal collection accuracy and low data transmission efficiency. Summary of the Invention

[0003] This application provides an energy consumption collection, monitoring and analysis cloud platform based on the Internet of Things, which is used to improve the accuracy and processing efficiency of energy consumption data collection. This application provides an energy consumption collection, monitoring and analysis cloud platform based on the Internet of Things, which includes:

[0004] The sensing unit is used to measure in real time and draw a time series curve graph based on the energy consumption data of the equipment, sample and obtain multiple target data based on the curve graph, and divide the data into groups;

[0005] an edge computing unit, configured to determine whether noise exists in the three adjacent target data in each data group according to the extreme value judgment results and the ratio judgment results corresponding to the three adjacent target data, and perform correction when noise exists to obtain corrected data;

[0006] an identification unit, configured to perform a fast frequency domain transformation based on the corrected data group to obtain a transformation result, and further analyze a fault state of the device based on the transformation result;

[0007] a sending unit, configured to compress the transformation result and send it to a cloud server when the fault state of the device is normal;

[0008] The cloud server is used to receive and perform inverse transformation on the transformation result, obtain and analyze the recovered energy consumption data, obtain energy consumption analysis results, and feed back to the device control module based on the energy consumption analysis results.

[0009] As a preferred technical solution of the present invention, the sensor module is configured as follows:

[0010] According to the measurement module in the sensing unit, multiple energy consumption data of the device are obtained in chronological order, a timing curve diagram of the multiple energy consumption data is drawn in a coordinate system according to the acquisition time, and sampling is performed according to a preset sampling period based on a reference point of the timing curve diagram as a starting position, and multiple target data are obtained. The target data within N energy consumption periods are taken as a data group, the ratio of the period of the energy consumption data to the preset period is a positive integer, and the energy consumption data includes at least device current data and voltage data.

[0011] As a preferred technical solution of the present invention, the edge computing unit is configured as follows:

[0012] sequentially calculating a first difference between the second data and the first data and a first ratio between the second data and the first data in every three adjacent target data, and further calculating a second difference between the third data and the second data and a second ratio between the third data and the second data;

[0013] An extreme value judgment result is obtained based on the first difference, the second difference and the extreme value reference position. When the extreme value judgment result is correct, a ratio judgment result is obtained based on the first ratio, the second ratio and the corresponding standard ratio. Whether there is noise in the three adjacent target data is judged based on the extreme value judgment result and the ratio judgment result. When no noise exists, the three adjacent target data are saved to a storage unit.

[0014] As a preferred technical solution of the present invention, the edge computing unit is further configured to: when the extreme value judgment result is normal and the ratio judgment result is abnormal, calculate the third ratio of the third data to the corresponding reference data, and also calculate the fourth ratio of any target data among the next three adjacent noise-free target data closest to the target data to the corresponding reference data. When the third ratio and the fourth ratio are consistent, there is no noise in the three adjacent target data; otherwise, there is noise.

[0015] As a preferred technical solution of the present invention, the edge computing unit is further configured as follows:

[0016] When there is noise in the three adjacent target data and the extreme value judgment result is abnormal, the three adjacent target data are used as the data to be corrected, and the target data corresponding to the positions of the three adjacent target data in the previous noise-free energy consumption cycle are obtained as the correction reference data. According to the ratio of the noise-free target data closest to the data to be corrected to the corresponding reference target data in the energy consumption cycle where the correction reference data is located, the product of the correction reference data and the ratio is used as the correction data to replace the data to be corrected.

[0017] As a preferred technical solution of the present invention, the identification unit is configured as follows:

[0018] A fast frequency domain transformation is performed based on the uncorrected target data corresponding to the corrected data group and the corrected data obtained after correction, and a transformation result is obtained. The transformation result is compared with the fault frequency domain characteristics to determine whether the equipment has a fault, and when a fault exists, an early warning message is sent to the management unit.

[0019] As a preferred technical solution of the present invention, the sending unit is used to compress the transformation result corresponding to the data group when the fault state of the device is normal, and send it to the cloud server at a set communication protocol and transmission rate.

[0020] As a preferred technical solution of the present invention, the fast frequency domain transform is a fast Fourier transform, and the inverse transform is an inverse fast Fourier transform.

[0021] The present invention also provides an energy consumption collection method based on the Internet of Things, which is implemented through the above-mentioned platform, and the method includes:

[0022] Step S1: Real-time measurement and drawing a time series curve graph based on the energy consumption data of the device, sampling based on the curve graph to obtain multiple target data, and dividing the data into groups;

[0023] Step S2: determining whether noise exists in the three adjacent target data according to the extreme value judgment results and the ratio judgment results corresponding to the three adjacent target data in each data group, and performing correction if noise exists to obtain corrected data;

[0024] Step S3: performing a fast frequency domain transformation based on the corrected data set to obtain a transformation result, and analyzing a fault state of the device based on the transformation result;

[0025] Step S4: When the fault state of the device is normal, compress the transformation result and send it to the cloud server;

[0026] Step S5: receiving and performing inverse transformation on the transformation result, acquiring and analyzing the restored energy consumption data, acquiring energy consumption analysis results, and feeding back the energy consumption analysis results to the device control module.

[0027] The present invention also provides a computer-readable storage medium having instructions stored thereon, and the above-mentioned method is implemented when the instructions are executed by a processor.

[0028] Beneficial effects: The present invention proposes an energy consumption collection, detection and analysis cloud platform based on the Internet of Things, which significantly improves the accuracy and processing efficiency of energy consumption data collection. Traditional energy consumption monitoring systems generally have problems such as insufficient sampling accuracy, serious noise interference, low data processing speed and poor transmission efficiency, which makes it difficult to meet the refined needs of modern industrial equipment for high-frequency energy consumption monitoring. The present invention integrates the sensing unit, edge computing unit, identification unit, sending unit and cloud server to form a closed-loop, high-efficiency data processing system, which solves the above problems. First, the sensing unit obtains high-frequency energy consumption data such as current and voltage in real time, and draws a timing curve diagram to realize the intuitive presentation of the working status of the equipment; secondly, the edge computing unit calculates the difference and ratio of three adjacent target data. The algorithm, combined with extreme value judgment and standard comparison, can effectively identify and correct noise interference in the data, improving data quality from the source. Furthermore, the recognition unit performs a fast Fourier transform based on the corrected data, converting the time domain signal into frequency domain features, enhancing the ability to identify the equipment's operating status, especially fault characteristics. When the system determines that the equipment is operating normally, the sending unit uploads the compressed frequency domain results to the cloud, significantly reducing the data transmission volume and improving network utilization. Finally, the cloud server restores the original data through inverse transformation and performs comprehensive analysis to provide energy consumption optimization suggestions and fault warnings to the equipment control module.

[0029] Through the mutual coordination of the above steps, the platform takes into account the comprehensive performance of high sampling accuracy, data purification processing, intelligent analysis and judgment, and efficient transmission response, truly realizing the energy consumption management closed loop of "edge perception + cloud intelligence", and providing reliable data support and energy-saving decision-making basis for industrial automation. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0031] Figure 1 This is a structural diagram of the energy consumption collection, monitoring and analysis cloud platform based on the Internet of Things in an embodiment of the present application;

[0032] Figure 2 This is a time series curve diagram of energy consumption data in an embodiment of the present application;

[0033] Figure 3 This is a flow chart of the energy consumption collection method based on the Internet of Things in an embodiment of the present application. DETAILED DESCRIPTION

[0034] The embodiment of the present application provides an energy consumption collection, monitoring and analysis cloud platform based on the Internet of Things. The terms "first", "second", "third", "fourth", etc. (if any) in the specification and claims of this application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "including" or "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0035] For ease of understanding, the specific process of the embodiment of the present application is described below. Figure 1 As shown, the energy consumption collection, monitoring and analysis cloud platform based on the Internet of Things in the embodiment of the present application includes:

[0036] The sensing unit is used to measure in real time and draw a time series curve graph based on the energy consumption data of the equipment, sample and obtain multiple target data based on the curve graph, and divide the data into groups;

[0037] Specifically, there are multiple production equipment in the factory. During the operation of the production equipment, the voltage sensor and the current sensor in the sensing unit collect the energy consumption data of the equipment in real time. The above energy consumption data includes at least current data and voltage data. The collection frequency of the above energy consumption data is relatively high. Therefore, Figure 2 As shown, the above-mentioned timing curve diagram with good continuity, i.e., a waveform diagram, can be drawn. The above-mentioned timing curve diagram has strong visualization and trend representation capabilities. Based on the above-mentioned timing curve diagram, with the relative zero point of the timing curve diagram as the reference starting position, i.e., the starting point of each cycle waveform, the energy consumption data is periodically sampled according to the preset sampling period to obtain a series of target data points, and these target data are divided into multiple data groups according to time continuity. Each data group contains data points within several sampling periods to support the subsequent noise identification and analysis process. The above-mentioned technical solution constructs the original data set of energy consumption data, providing a structured foundation for subsequent algorithm processing. Through high-frequency acquisition and reasonable period division, it can more comprehensively capture the subtle fluctuations in energy consumption during equipment operation, facilitate subsequent noise removal and accurate frequency domain analysis, and lay the foundation for subsequent denoising and correction of the above-mentioned energy consumption data.

[0038] an edge computing unit, sequentially judging whether noise exists in the three adjacent target data according to extreme value judgment results and ratio judgment results corresponding to the three adjacent target data in the data group, and correcting the noise when it exists to obtain corrected data;

[0039] Specifically, in each data group, the edge computing unit will extract three adjacent target data points in turn. The three adjacent target data can be obtained by means of a sliding window. By calculating their difference (first difference, second difference) and ratio (first ratio, second ratio), it is used to determine whether the data segment contains noise. Then, by using the method of extreme value position judgment and standard ratio comparison, it is determined whether the three-point data is reasonable. For example, if the middle point is an extreme value (such as a peak or a trough) and deviates greatly from the preset extreme value position, it means that the point may be interfered by noise. If noise is detected, then Find noise-free data in the corresponding reference period, and calculate a set of correction data based on the position correspondence principle and amplitude ratio to replace the original abnormal data. The above technical solution can effectively eliminate sampling errors caused by factors such as sudden start and stop of equipment, electromagnetic interference or sensor fluctuations, and improve data accuracy and stability. This is especially true in industrial environments where power fluctuations are frequent and complex, and missampling occurs from time to time. Therefore, the noise identification and correction mechanism not only improves the credibility of subsequent analysis, but also ensures the accuracy of the entire energy consumption data chain, avoiding the risk of misjudging equipment abnormalities or missed detections.

[0040] an identification unit, configured to perform a fast frequency domain transformation based on the corrected data group to obtain a transformation result, and further analyze a fault state of the device based on the transformation result;

[0041] Specifically, after the data is cleaned and corrected, the identification unit performs a rapid frequency domain transform on each set of target data, specifically using Fast Fourier Transform (FFT) technology. FFT converts time-domain energy consumption data into frequency-domain data, revealing its energy distribution across frequency components. It also reduces the amount of data generated by the transform, facilitating improved transmission efficiency. This frequency-domain representation is more suitable for capturing the periodic characteristics of equipment operation. The system then compares the transformed spectrum with preset fault feature templates (such as specific frequency peaks and harmonic enhancements) to determine whether there are abnormal equipment operations or fault risks. This technical solution enables a higher level of intelligent recognition of energy consumption data. The spectral perspective provided by FFT technology makes it easier to identify "invisible faults" in equipment operation, such as abnormal motor vibration, partial discharge, or periodic fluctuations caused by aging. This significantly improves the accuracy of fault identification and early warning capabilities, allowing companies to intervene and repair problems before they worsen, saving costs and reducing risks.

[0042] a sending unit, configured to compress the frequency domain transformation result and send it to a cloud server when the fault state of the device is normal;

[0043] Specifically, after confirming that the device is operating normally, the system initiates the data transmission process. To improve communication efficiency, the sending unit compresses the frequency domain transform results to minimize data footprint. After compression, the data is sent to the cloud server via a predefined communication protocol (such as MQTT, TCP / IP, etc.) and transmission rate (such as LTE, NB-IoT). Data is uploaded only when the fault status is normal. This prevents abnormal data from disrupting the server processing flow and reduces bandwidth waste. This technical solution significantly improves the data transmission efficiency of the entire system and is particularly suitable for edge computing environments and low-bandwidth IoT scenarios, such as factory floors, basements, and remote mountainous areas with poor network coverage. By determining device status before sending data, system resource utilization is further improved, ineffective communication and cloud load are reduced, and efficient system operation and scalability are ensured.

[0044] The cloud server is used to receive and perform inverse transformation on the frequency domain transformation result, obtain and analyze the recovered energy consumption data, obtain energy consumption analysis results, and feed back to the device control module based on the energy consumption analysis results.

[0045] Specifically, after receiving the compressed frequency-domain data, the cloud server uses an inverse fast Fourier transform to restore it to time-domain data. This reconstructed energy consumption data is further analyzed for trend assessment, equipment condition determination, and energy efficiency analysis. The final analysis results are fed back to the equipment control module, which can be used to automatically adjust operating strategies (such as reducing power and adjusting energy-saving modes) or to send warnings to users for manual intervention. Through the interaction of these technical solutions, a closed-loop energy consumption monitoring and feedback system has been established. This system eliminates isolation in the operating status of equipment and enables dynamic interaction with the cloud data center. Through cloud computing-based data mining and historical modeling, the system not only performs real-time analysis but also identifies long-term abnormal energy consumption trends, enabling dynamic adjustment of energy-saving strategies and predictive maintenance, thereby promoting the transformation of industrial equipment management towards intelligent, data-driven management.

[0046] Furthermore, the sensing unit is configured as follows:

[0047] According to the measurement module in the sensing unit, multiple energy consumption data of the device are obtained in chronological order, a timing curve diagram of the multiple energy consumption data is drawn in a coordinate system according to the acquisition time, and sampling is performed according to a preset sampling period based on a reference point of the timing curve diagram as a starting position, and multiple target data are obtained. The target data within N energy consumption periods collected in real time are used as a data group, the ratio of the period of the energy consumption data to the preset sampling period is a positive integer, and the energy consumption data includes at least device current data and voltage data.

[0048] Specifically, the measurement module in the sensing unit acquires multiple energy consumption data of the device in chronological order. These data include current data and voltage data of the device, etc. Based on the collected energy consumption data, the sensing unit draws a timing curve graph of these data in the coordinate system according to the collection time. The timing curve graph, as a visual representation of the data, can more clearly understand the energy consumption fluctuations of the device in different time periods, and provide clear basic data for subsequent analysis. Using the drawn timing curve graph, the system samples the data according to the preset sampling period. Since the period of the above energy consumption data is an integer multiple of the above preset sampling period, it can be guaranteed that there can be an integer number of sampling points, namely the above target data, in each period of the above energy consumption data, thereby facilitating sampling and subsequent processing, and dividing all the above target data within N energy consumption periods into a data group. The above technical solution can not only accurately obtain the above target data, but also facilitate noise detection and correction of subsequent target data through grouping.

[0049] Furthermore, the edge computing unit is configured as follows:

[0050] sequentially calculating a first difference between the second data and the first data and a first ratio between the second data and the first data in every three adjacent target data, and further calculating a second difference between the third data and the second data and a second ratio between the third data and the second data;

[0051] An extreme value judgment result is obtained based on the first difference, the second difference and the extreme value position reference point. When the extreme value judgment result is correct, a ratio judgment result is obtained based on the first ratio, the second ratio and the corresponding standard. Whether there is noise in the three adjacent target data is judged based on the extreme value judgment result and the ratio judgment result. When no noise exists, the three adjacent target data are saved to a storage unit.

[0052] Specifically, the edge computing unit first processes the three adjacent target data in sequence. The three target data are the first data, the second data and the third data in the order of acquisition, and calculates the difference between the adjacent data. Specifically, the first difference between the second data and the first data, the second difference between the third data and the second data, and the ratio between the adjacent data are calculated. The first ratio between the second data and the first data, and the second ratio between the third data and the second data are calculated. The edge computing unit uses the first difference, the second difference, the first ratio and the second ratio to make extreme value judgments, and judges whether the data is abnormal by referring to the standard of the extreme value position. Since the first data, the second data and the third data are sorted according to the acquisition time, when the first difference is a positive number greater than 0 and the second difference is a negative number less than 0, the second data is the maximum extreme value point. When the first difference is a negative number less than 0 and the second difference is a positive number greater than 0, the second data position is the minimum extreme value point. Conversely, when the first difference and the second difference are both positive or negative, the second data position is the minimum extreme value point. The data position is not an extreme point, and the second data position is compared with the corresponding extreme value reference position. When the distance between the second data position and the extreme value reference position is less than or equal to the set distance, and the extreme value type is different (for example, the maximum extreme value point at the second data position, but the extreme value reference point is the minimum extreme value point), or the distance between the second data position and the extreme value reference position is greater than the set distance, and it is an extreme point, that is, the extreme point is detected at the position of the extreme point, indicating that the extreme value judgment result is abnormal, that is, there is noise in the first data to the third data including the third data position, for example, Figure 2In the first cycle, that is, there is a sawtooth wave interference position between 0-t1. On the contrary, when the distance between the second data position and the extreme value reference position is less than or equal to the set distance, and the extreme value type is the same, or the distance between the second data position and the extreme value reference position is greater than the set distance, and it is not an extreme value point, it means that the extreme value position judgment result is correct. However, the correct extreme value position judgment result does not mean that there is no noise between the first data and the third data, including the third data position. The edge computing unit further judges whether there is noise with the same amplitude change trend as that between the first data and the third data according to the calculated ratio judgment result and the corresponding standard, and compares the first ratio between the second data and the first data and the second ratio between the third data and the second data with the corresponding first standard ratio and the second standard ratio respectively. In the first When the ratio is consistent with the first standard ratio and the second ratio is also consistent with the second standard ratio, there is no noise between the above-mentioned first data and the above-mentioned third data, otherwise there is noise, wherein the first reference data, the second reference data and the third reference data are obtained according to the positions corresponding to the above-mentioned first data, the second data and the third data in the reference period of the above-mentioned energy consumption data, and the ratio of the above-mentioned second reference data to the above-mentioned first reference data is used as the above-mentioned first standard ratio, and the ratio between the above-mentioned third reference data and the above-mentioned second reference data is used as the above-mentioned second standard ratio. The reference period of the above-mentioned energy consumption data provides a benchmark for noise judgment, and when there is no noise in the above-mentioned three adjacent target data, the above-mentioned three adjacent target data are saved in the above-mentioned storage unit. The above-mentioned technical solution can not only accurately judge whether there is noise in the above-mentioned data group, but also provide a basis for correcting the corresponding target data in other subsequent energy consumption periods.

[0053] Furthermore, the edge computing unit is also configured to: when the extreme value judgment result is normal and the ratio judgment result is abnormal, calculate the third ratio of the third data to the corresponding reference data, and also calculate the fourth ratio of any target data among the next three adjacent noise-free target data closest to the target data to the corresponding reference data. When the third ratio and the fourth ratio are consistent, there is no noise in the three adjacent target data; otherwise, there is noise.

[0054] Specifically, when the extreme value judgment result is normal and the ratio judgment result is abnormal, there is also a special case. Since the third data is a new data added after the first data is removed compared with the previous three adjacent target data, the position of the third data may be a critical point of amplitude change, which causes the second ratio to be inconsistent with the corresponding second standard ratio, thereby making the ratio judgment result abnormal. At this time, it is necessary to calculate the third ratio of the third data and the corresponding reference data. The corresponding reference data of the third data is the energy consumption data corresponding to the third data position in the reference period of the energy consumption data. Once the subsequent target data corresponding position When there is noise at the position, the corresponding amplitude may be inaccurate and have no reference value. Therefore, the third ratio is compared with the fourth ratio of the target data at the nearest and noise-free position and the corresponding reference data. When the third ratio is consistent with the fourth ratio, it means that the third data position is a negative value change critical point. At this time, it can be judged that there is no noise at the third data position. Otherwise, there is noise at the third data position. The technical solution can further accurately determine whether there is noise in the three adjacent target data when the extreme value judgment result is normal and the ratio judgment result is abnormal, thereby ensuring the accuracy of the target data.

[0055] Furthermore, the edge computing unit is further configured as:

[0056] When there is noise in the three adjacent target data and the extreme value judgment result is abnormal, the three adjacent target data are used as the data to be corrected, and the target data corresponding to the positions of the three adjacent target data in the previous noise-free energy consumption cycle are obtained as the correction reference data. According to the ratio of the noise-free target data closest to the data to be corrected to the corresponding reference target data in the energy consumption cycle where the correction reference data is located, the product of the correction reference data and the ratio is used as the correction data to replace the data to be corrected.

[0057] Specifically, when there is noise in the above-mentioned three adjacent target data, the above-mentioned three adjacent target data are used as the above-mentioned data to be corrected. Since the target data corresponding to the position of the above-mentioned data to be corrected in the previous noise-free energy consumption cycle of the energy consumption cycle corresponding to the above-mentioned data to be corrected is used as the above-mentioned correction reference data, the above-mentioned correction reference data reflects the energy consumption characteristics of the current device, that is, the waveform characteristics. Therefore, the waveform characteristics of the above-mentioned correction reference data can be used as the reference waveform of the above-mentioned data to be corrected. However, since the amplitude characteristics of the above-mentioned data to be corrected and the above-mentioned correction reference data are not necessarily the same, the ratio of the above-mentioned noise-free target data closest to the above-mentioned data to be corrected to the above-mentioned reference target data corresponding to the above-mentioned energy consumption cycle where the above-mentioned correction reference data is located is also used. Among them, the above-mentioned noise-free target data and the above-mentioned reference target data have the same position in the corresponding energy consumption cycle. The above-mentioned ratio reflects the amplitude ratio of the above-mentioned data to be corrected relative to the above-mentioned correction reference data. Therefore, the product of the above-mentioned correction reference data and the above-mentioned ratio is used as the corrected data after the above-mentioned data to be corrected is corrected. The above-mentioned technical solution can accurately correct the above-mentioned data to be corrected, thereby improving the collection accuracy of energy consumption data.

[0058] Furthermore, the identification unit is configured as follows:

[0059] A fast frequency domain transformation is performed based on the uncorrected target data corresponding to the corrected data group and the corrected data obtained after correction, and a transformation result is obtained. The transformation result is compared with the fault frequency domain characteristics to determine whether the equipment has a fault, and when a fault exists, an early warning message is sent to the management unit.

[0060] Specifically, the corrected data group undergoes a fast frequency domain transformation. Frequency domain transformation can convert time domain signals into frequency domain signals, such as fast Fourier transform, so that periodic and regular information can be clearly displayed. Compared with time domain data, frequency domain data is easier to identify the frequency components in the operation of the equipment, which helps to discover potential fault signals. The transformed frequency domain data is compared with the preset fault frequency domain characteristics. Based on the comparison between the transformation results and the fault characteristics, the system can determine whether the equipment has a fault. If the frequency domain characteristics of the equipment are detected to match the known fault mode, the system will determine that the equipment has a fault and send the fault information to the management unit when the above equipment has a fault. The above technical solution can make a fault judgment based on the frequency domain transformation results corresponding to the corrected data group and report the fault information in a timely manner, thereby reminding the user in time.

[0061] Furthermore, the sending unit is used to compress the transformation result corresponding to the data group and send it to the cloud server at a set communication protocol and transmission rate when the fault state of the device is normal.

[0062] Specifically, when the fault state of the above-mentioned device is normal, that is, when the device is operating normally, the above-mentioned transformation results corresponding to the above-mentioned data group are compressed, so that the amount of energy consumption data is greatly reduced, which helps to alleviate the bandwidth pressure of the low-speed communication channel, so that even in the environment of bandwidth limitation and low-speed transmission, the system can still transmit data stably, ensuring real-time performance.

[0063] Furthermore, the fast frequency domain transform is a fast Fourier transform, and the inverse transform is an inverse fast Fourier transform.

[0064] The present invention also provides an energy consumption collection method based on the Internet of Things, which is implemented through the above platform, such as Figure 3 As shown, the method includes:

[0065] Step S1: Real-time measurement and drawing a time series curve graph based on the energy consumption data of the device, sampling and obtaining multiple target data based on the curve graph, dividing the data into groups, and determining whether there is an abnormality in the data group through the recognition unit;

[0066] Step S2: determining whether noise exists in the three adjacent target data according to the extreme value judgment results and the ratio judgment results corresponding to the three adjacent target data in each data group, and performing correction if noise exists to obtain corrected data;

[0067] Step S3: performing a fast frequency domain transformation based on the corrected data set to obtain a transformation result, and analyzing a fault state of the device based on the transformation result;

[0068] Step S4: When the fault state of the device is normal, compress the transformation result and send it to the cloud server;

[0069] Step S5: receiving and performing inverse transformation on the transformation result, acquiring and analyzing the restored energy consumption data, acquiring energy consumption analysis results, and feeding back the energy consumption analysis results to the device control module.

[0070] The present invention also provides a computer-readable storage medium having instructions stored thereon, and the above-mentioned method is implemented when the instructions are executed by a processor.

[0071] In summary, the present invention forms a closed-loop, high-efficiency data processing system by integrating the sensing unit, edge computing unit, identification unit, sending unit and cloud server, which solves the above problems. First, the sensing unit obtains high-frequency energy consumption data such as current and voltage in real time, and draws a timing curve diagram, thereby realizing an intuitive presentation of the working status of the equipment; secondly, the edge computing unit calculates the difference and ratio of three adjacent target data, and combines extreme value judgment with standard comparison to effectively identify and correct noise interference in the data, thereby improving data quality from the source; further, the identification unit performs fast Fourier transform based on the corrected data, converts the time domain signal into frequency domain features, and enhances the accuracy of the data. It strengthens the ability to identify the operating status of equipment, especially fault characteristics; when the system determines that the equipment is operating normally, the sending unit uploads the compressed frequency domain results to the cloud, which greatly reduces the data transmission volume and improves network utilization; finally, the cloud server restores the original data through inverse transformation and performs comprehensive analysis to provide energy consumption optimization suggestions and fault warnings for the equipment control module; through the mutual cooperation between the above technical solutions, the platform takes into account the comprehensive performance of high sampling accuracy, data purification processing, intelligent analysis and judgment, and efficient transmission response, truly realizing the energy consumption management closed loop of "edge perception + cloud intelligence", and providing reliable data support and energy-saving decision-making basis for industrial automation.

[0072] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described systems, systems and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0073] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk, and other media that can store program code.

[0074] As described above, the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the above embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the above embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. An energy consumption collection, monitoring and analysis cloud platform based on the Internet of Things, characterized by: The platform includes: The sensing unit is used to measure in real time and draw a time series curve graph based on the energy consumption data of the equipment, sample and obtain multiple target data based on the curve graph, and divide the data into groups; an edge computing unit, configured to determine whether noise exists in the three adjacent target data in the data group according to the extreme value judgment results and the ratio judgment results corresponding to the three adjacent target data in the data group, and perform correction when noise exists to obtain corrected data; an identification unit, configured to perform a rapid frequency domain transformation based on the target data and the correction data to obtain a frequency domain transformation result, and further analyze a fault state of the device based on the frequency domain transformation result; a sending unit, configured to compress the frequency domain transformation result and send it to a cloud server when the fault state of the device is normal; The cloud server is configured to receive and perform an inverse transformation on the frequency domain transformation result, obtain and analyze the recovered energy consumption data, obtain an energy consumption analysis result, and feed back the energy consumption analysis result to the device control module; The edge computing unit is further configured as follows: sequentially calculating a first difference between the second data and the first data and a first ratio between the second data and the first data in every three adjacent target data, and further calculating a second difference between the third data and the second data and a second ratio between the third data and the second data; obtaining an extreme value judgment result based on the first difference, the second difference, and the extreme value reference position; obtaining a ratio judgment result based on the first ratio, the second ratio, and the corresponding standard ratio when the extreme value judgment result is correct; judging whether there is noise in the three adjacent target data based on the extreme value judgment result and the ratio judgment result; and storing the three adjacent target data in a storage unit when no noise is present; When the extreme value judgment result is normal and the ratio judgment result is abnormal, the third ratio of the third data to the corresponding reference data is calculated, and the fourth ratio of any target data among the next three adjacent noise-free target data closest to the target data to the corresponding reference data is calculated. When the third ratio and the fourth ratio are consistent, there is no noise in the three adjacent target data; otherwise, there is noise.

2. The platform according to claim 1, characterized in that The sensing unit is configured as follows: According to the measurement module in the sensing unit, multiple energy consumption data of the device are obtained in chronological order, a timing curve diagram of the multiple energy consumption data is drawn in a coordinate system according to the acquisition time, and sampling is performed according to a preset sampling period based on a reference point of the timing curve diagram as a starting position, and multiple target data are obtained. The target data within N energy consumption periods are taken as a data group, the ratio of the period of the energy consumption data to the preset sampling period is a positive integer, and the energy consumption data includes at least device current data and voltage data.

3. The platform according to claim 1, characterized in that The edge computing unit is also configured to: When there is noise in the three adjacent target data and the extreme value judgment result is abnormal, the three adjacent target data are used as the data to be corrected, and the target data corresponding to the positions of the three adjacent target data in the previous noise-free energy consumption cycle are obtained as the correction reference data. According to the ratio of the noise-free target data closest to the data to be corrected to the corresponding reference target data in the energy consumption cycle where the correction reference data is located, the product of the correction reference data and the ratio is used as the correction data to replace the data to be corrected.

4. The platform according to claim 1, characterized in that The identification unit is configured as follows: A fast frequency domain transformation is performed based on the uncorrected target data corresponding to the corrected data group and the corrected data obtained after correction, and a transformation result is obtained. The transformation result is compared with the fault frequency domain characteristics to determine whether the equipment has a fault, and when a fault exists, an early warning message is sent to the management unit.

5. The platform according to claim 1, characterized in that The sending unit is used to compress the transformation result corresponding to the data group and send it to the cloud server according to the set communication protocol and transmission rate when the fault state of the device is normal.

6. The platform according to claim 1, characterized in that The fast frequency domain transform is a fast Fourier transform, and the inverse transform is an inverse fast Fourier transform.

7. A method for collecting energy consumption based on the Internet of Things, implemented by the platform according to any one of claims 1 to 6, characterized in that: The method comprises: Step S1: Real-time measurement and drawing a time series curve graph based on the energy consumption data of the device, sampling and obtaining multiple target data based on the curve graph, dividing the data into groups, and determining whether there is an abnormality in the data group through the recognition unit; Step S2: sequentially judging the extreme value judgment results and ratio judgment results corresponding to three adjacent target data in each data group to judge whether there is noise in the three adjacent target data, and correcting the noise if there is noise to obtain corrected data; Step S3: performing a fast frequency domain transformation based on the corrected data set to obtain a transformation result, and analyzing a fault state of the device based on the transformation result; Step S4: When the fault state of the device is normal, compress the transformation result and send it to the cloud server; Step S5: receiving and performing inverse transformation on the transformation result, acquiring and analyzing the restored energy consumption data, acquiring energy consumption analysis results, and feeding back the energy consumption analysis results to the device control module.

8. A computer-readable storage medium having instructions stored thereon, characterized in that: When the instructions are executed by a processor, the method according to claim 7 is implemented.

Images