Electromechanical equipment optimization control system based on edge calculation

Abstract

The invention relates to the technical field of electromechanical equipment monitoring, in particular to an electromechanical equipment optimization control system based on edge computing. According to the invention, the state of the electromechanical equipment is accurately monitored through real-time data acquisition and edge calculation processing, and fault diagnosis is carried out. In combination with the first support vector machine model and the second support vector machine model, fault risks are recognized in advance through equipment state data and vibration frequency energy, maintenance or monitoring signals are output through signal logic matching, effective maintenance of equipment is ensured, and shutdown and damage are avoided. Through an intelligent algorithm and signal matching, the system realizes intelligent equipment management, automatically notifies maintenance of equipment needing maintenance, ensures normal operation of the equipment, avoids unnecessary operation, improves the resource utilization efficiency, optimizes the management process, reduces manual intervention, and improves the management automation and efficiency.

Description

Technical Field

[0001] The present invention relates to the technical field of electromechanical equipment monitoring, and specifically to an electromechanical equipment optimization control system based on edge computing. Background Art

[0002] With the continuous improvement of industrial automation, the role of electromechanical equipment in production and operation is becoming increasingly important. The normal operation of equipment is directly related to production efficiency and product quality. Therefore, how to achieve intelligent management of electromechanical equipment and ensure its efficient, stable, and reliable operation has become a pressing issue in the current industrial field.

[0003] Existing technologies primarily monitor electromechanical equipment through traditional manual inspections and scheduled maintenance. This approach is not only inefficient but also susceptible to human error. It fails to detect potential equipment failures in real time, which can easily lead to equipment failures, production downtime, and increased repair costs. Furthermore, existing equipment fault diagnosis methods often rely on simple rules or manual judgment, lacking in-depth analysis and intelligent prediction of equipment operating data, making it difficult to effectively predict the remaining life of equipment and failure trends.

[0004] In response to the above problems, it is necessary to propose an optimization control system for electromechanical equipment based on edge computing. Summary of the Invention

[0005] The purpose of the present invention is to solve the problems existing in the background technology and propose an electromechanical equipment optimization control system based on edge computing.

[0006] The purpose of the present invention can be achieved through the following technical solutions:

[0007] The present invention provides an electromechanical equipment optimization control system based on edge computing, including a data acquisition module, an edge data processing and optimization module, a central control signal matching and scheduling module, and an execution and feedback module.

[0008] The data acquisition module is responsible for collecting sensor signals from electromechanical devices in real time and converting them into status data for each device, including ambient temperature, humidity, power, current, voltage, and vibration. Additional data collection is performed based on the device type. The collected data from different electromechanical devices is packaged according to a specific transmission protocol and sent to the edge data processing and optimization module via the communication network.

[0009] The status data of electromechanical equipment collected includes:

[0010] Ambient temperature T(i, t), ambient humidity H(i, t), apparent power S(i, t), effective power P(i, t), effective current I(i, t), effective voltage V(i, t) and vibration frequency a(i, t); where i is the device serial number of all monitored electromechanical devices, i=1, 2, ..., n; n is the total number of all monitored electromechanical devices; and t is the time of data collection.

[0011] Additional data collection is performed based on the device type. The specific process of packaging the data collected from different devices according to a specific transmission protocol is as follows:

[0012] Get the device type of each electromechanical device i, including motor, generator, compressor, fan, and transformer.

[0013] For electromechanical devices with motors, additional status data is collected: the operating mode identifier C1(i, t) and the load status identifier C2(i, t). This data is organized in JSON format. The MQTT protocol is used for data transmission, and all collected status data for these electromechanical devices is sent to the edge data processing and optimization module. Operating modes include: running, standby, and stopped; load status identifiers include: abnormal load and normal load. The operating mode identifier C1(i, t) uses 1, 0, and -1 to represent running, standby, and stopped, respectively; the load status identifier C2(i, t) uses 0 and 1 to represent abnormal load and normal load, respectively.

[0014] For electromechanical devices with the device type of generator, additional status data is obtained: input power P_in(i, t) and output frequency P_out(i, t), and this status data is organized in JSON format. Using the AMQP protocol for data transmission, all collected status data for electromechanical devices with the device type of motor is sent to the edge data processing and optimization module.

[0015] For compressor-type electromechanical devices, additional status data is collected: pressure F(i, t), flow rate Q(i, t), and compression efficiency η(i, t). This data is organized in JSON format. MQTT is used for data transmission, and all collected status data for compressor-type electromechanical devices is sent to the edge data processing and optimization module.

[0016] For electromechanical devices of the wind turbine type, additional status data is collected: speed ω(i, t) and output power P_out(i, t). This data is organized in JSON format. HTTP is used for data transmission, and all collected status data for electromechanical devices of the wind turbine type is sent to the edge data processing and optimization module.

[0017] For electromechanical devices with transformers, additional status data is collected: input voltage Vi_in(i, t), output voltage V_out(i, t), and load factor L(i, t). This data is organized in JSON format. Using AMQP for data transmission, all collected status data for electromechanical devices with transformers is sent to the edge data processing and optimization module.

[0018] The edge data processing and optimization module is responsible for receiving status data from the data acquisition module, performing preliminary real-time processing of the data according to the preset control algorithm, and performing feature extraction and performance evaluation on the collected status data of each electromechanical device.

[0019] The specific process of feature extraction of the collected status data of each electromechanical device is as follows:

[0020] Extract key features from the status data of each electromechanical device, including vibration frequency spectrum features, current harmonic features, and power factor features.

[0021] The specific process of vibration frequency spectrum feature extraction is:

[0022] Input the vibration frequency a(i, t) of each electromechanical device i into the formula: The vibration frequency of each electromechanical device is converted from a time domain signal to a frequency domain signal through fast Fourier transform. Here, f is the frequency obtained after Fourier transform, X(f) is the transformed frequency domain signal, t is time, and j is the imaginary unit.

[0023] The spectrum features of the transformed frequency domain signal X(f) are extracted to obtain the fundamental frequency f0, peak frequency fmax and frequency components f1, f2, f3, ..., fmax.

[0024] Extract the energy E(f0) of the fundamental frequency f0 and the energies E(f1), E(f2), E(f3), ..., E(fmax) of the frequency components f1, f2, f3, ..., fmax. The energy calculation formula for the frequency components is:

[0025] ; Where m is the frequency component index, fm=f1,f2,f3,...,fmax;

[0026] The specific process of current harmonic feature extraction is:

[0027] Harmonic components are extracted from the effective current I(i, t) through Fourier transform, including the fundamental current I1(i) and higher harmonic currents Ik(i)=I2(i), I3(i), ..., IK(i); where k=1, 2, ..., K; K is the maximum order of the extracted harmonic current.

[0028] By preset formula Calculate the total harmonic distortion THD(i) of each electromechanical device.

[0029] The specific process of power factor feature extraction is:

[0030] Input the apparent power S(i, t) and effective power P(i, t) of each electromechanical device into the formula:

[0031] , calculate the power factor PF(i, t) of each electromechanical device i at time t.

[0032] By default formula:

[0033]

[0034] Calculate the first fluctuation coefficient of each electromechanical device i1 between the current time t and time tN and the second coefficient of fluctuation .

[0035] The first fluctuation coefficient represents the second-order moment of the power factor PF(i, t) within the past preset time, and the second fluctuation coefficient represents the maximum instantaneous rate of change of the power factor PF(i, t) within the past preset time. N is the preset power factor comparison monitoring period, and q is the timestamp index, ranging from the historical time t-N+1 to the current time t.

[0036] The central control signal matching and scheduling module is responsible for storing the first and second support vector machine models, which respectively detect abnormal conditions based on status data and extracted vibration frequency energy. Fault diagnosis and performance evaluation are performed based on the status data and extracted key features of the electromechanical devices. Signal logic matching is performed based on the calculation results of the edge data processing and optimization module to generate control signals uniquely corresponding to each electromechanical device to adjust its operating status.

[0037] The specific process of fault diagnosis and performance evaluation is as follows:

[0038] Obtain all state data of each electromechanical device i at time t to form a state data vector X1(i)={T(i, t), H(i, t), S(i, t), P(i, t), I(i, t), V(i, t), a(i, t), C1(i, t), C2(i, t), P_in(i, t), P_out(i, t), F(i, t), Q(i, t), η(i, t), ω(i, t), Vi_in(i, t), V_out(i, t), L(i, t)}; the value of the data not collected is recorded as 0.

[0039] Obtain the frequency energy E(f1), E(f2), E(f3), ..., E(fmax) of the vibration components extracted from each electromechanical device i, and form a vibration frequency energy vector X2(i) = {E(f1), E(f2), E(f3), ..., E(fmax)};

[0040] Input the state data vector X1(i) into the first support vector machine model:

[0041] ;

[0042] Obtain the first judgment label Y1(i) representing the abnormal state of the surface data; where W1 and b1 are the decision hyperplane normal vector and weight vector of the first support vector machine, respectively, which are determined by the training data;

[0043] Input the vibration frequency energy vector X2(i) into the second support vector machine:

[0044]

[0045] Obtain the second judgment label Y2(i) representing the abnormal state of the vibration frequency; where W2 and b2 are the decision hyperplane normal vector and weight vector of the second support vector machine, respectively, which are determined by the training data;

[0046] Among them, the first judgment label Y1 (i) and the second judgment label Y2 (i) are variables with a value range of 1 to 100. The larger the value, the greater the probability of an abnormality, and the smaller the value, the closer the electromechanical equipment is to normal use.

[0047] The process of training the first support vector machine model and the second support vector machine to determine the specific values ​​of W1, b1, W2 and b2 is as follows:

[0048] Perform learning data sampling, mark a set of state data vectors X1(i) and vibration frequency energy vectors X2(i) from the same electromechanical device i, and preset their corresponding standard output values and If the electromechanical equipment i fails, and are all 0; otherwise, and Both are 1.

[0049] The state data vector X1(i) and the vibration frequency energy vector X2(i) obtained by sampling the learning data are input into the first support vector machine model and the second support vector machine to obtain the output first judgment label Y1(i) and the second judgment label Y2(i).

[0050] Preset the initial values ​​of w1, w2, b1 and b2, and define the optimization objective function: , perform joint optimization operations on w1, w2, b1 and b2.

[0051] C is the preset regularization parameter that controls the degree of penalty for misclassification. and are the first and second slack variables, respectively, representing the controlled distance between the data point and the decision boundary in the first support vector machine model and the second support vector machine.

[0052] The convex optimization problem contained in the optimization objective function is solved by the Lagrange multiplier method to obtain the specific values ​​of w1, w2, b1 and b2.

[0053] The final specific values ​​of w1, w2, b1 and b2 are brought back to the first support vector machine model and the second support vector machine to complete the final training.

[0054] The specific process of signal logic matching is:

[0055] A first judgment label Y1 (i) and a second judgment label Y2 (i) of each electromechanical device i are obtained.

[0056] Obtain the total harmonic distortion THD(i) of each electromechanical device i;

[0057] Get the first fluctuation coefficient of each electromechanical device i and the second coefficient of fluctuation ;

[0058] If the first judgment tag Y1(i) of the electromechanical device i is greater than the preset threshold, and the second judgment tag Y2(i) is greater than the preset threshold, and the first fluctuation coefficient is greater than the preset threshold, and the second fluctuation coefficient is greater than the preset threshold, If the total harmonic distortion THD(i) is greater than the preset threshold, then the maintenance signal of the electromechanical equipment i is output;

[0059] If the first judgment tag Y1(i) of the electromechanical device i is greater than the preset threshold, and the second judgment tag Y2(i) is greater than the preset threshold. At the same time, if the following conditions are met: the first fluctuation coefficient is greater than the preset threshold, or the second fluctuation coefficient is greater than the preset threshold, If the value of the total harmonic distortion THD(i) is greater than the preset threshold, or the total harmonic distortion THD(i) is greater than the preset threshold, a continuous monitoring signal of the electromechanical device i is output;

[0060] If the first determination tag Y1 (i) of the electromechanical device i is identified to be less than or equal to the preset threshold, and the second determination tag Y2 (i) is less than or equal to the preset threshold, a normal operation signal of the electromechanical device i is output.

[0061] The execution and feedback module executes specific maintenance optimization operations according to the control signals uniquely corresponding to each electromechanical device, and provides real-time feedback on the operating status of the electromechanical device during the execution of the maintenance optimization operation.

[0062] For electromechanical equipment i that generates maintenance signals, immediately notify the administrator to organize personnel to carry out repairs and equipment maintenance.

[0063] For electromechanical equipment i that generates continuous monitoring signals, the administrator is immediately notified to organize personnel to conduct manual evaluation of the operating status, cleaning, and fault inspection.

[0064] For electromechanical equipment i that generates normal operating signals, no unnecessary operations are taken.

[0065] Compared with the prior art, the present invention has the following beneficial effects:

[0066] 1. By collecting real-time data from electromechanical equipment and processing it through edge computing, the present invention can monitor the status of equipment in real time and diagnose faults. By introducing the first and second support vector machine models, combined with equipment status data and vibration frequency energy, it can accurately determine whether the equipment is experiencing anomalies, thereby identifying potential equipment failure risks in advance. The system determines whether the equipment requires maintenance or continuous monitoring through signal logic matching, and promptly outputs maintenance signals or continuous monitoring signals to ensure that the equipment is effectively maintained and inspected, avoiding downtime losses and equipment damage caused by equipment failures.

[0067] 2. This invention not only detects abnormal equipment conditions but also evaluates equipment performance. By continuously monitoring and comparing equipment operating status and failure modes, the system can assess equipment health and provide real-time feedback, assisting equipment managers in optimizing equipment operation strategies and extending equipment life. Furthermore, by monitoring parameters such as the fluctuation coefficient and total harmonic distortion (THD), the system can identify equipment operational stability and make timely adjustments to ensure long-term stable operation.

[0068] 3. This invention achieves intelligent equipment maintenance and management through intelligent algorithms and signal logic matching, combined with the equipment's operating status and fault diagnosis results. For equipment generating maintenance signals, the system automatically notifies the administrator to organize personnel for repair and maintenance operations; for equipment generating continuous monitoring signals, the system prompts the administrator to conduct manual evaluation and inspection. For equipment in normal operating status, the system avoids unnecessary operations, improves resource utilization, optimizes equipment management processes, and reduces the need for manual intervention, thereby enhancing the automation and efficiency of overall management. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] In order to facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings:

[0070] Figure 1 This is a system block diagram of the present invention. DETAILED DESCRIPTION

[0071] The technical solutions of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the embodiments described are only some of the embodiments of the present invention, not all of them. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0072] See also Figure 1 As shown in the figure, the electromechanical equipment optimization control system based on edge computing includes a data acquisition module, an edge data processing and optimization module, a central control signal matching and scheduling module, and an execution and feedback module.

[0073] The data acquisition module is responsible for collecting sensor signals from electromechanical devices in real time and converting them into status data for each device, including ambient temperature, humidity, power, current, voltage, and vibration. Additional data collection is performed based on the device type. The collected data from different electromechanical devices is packaged according to a specific transmission protocol and sent to the edge data processing and optimization module via the communication network.

[0074] The status data of electromechanical equipment collected includes:

[0075] Ambient temperature T(i, t), ambient humidity H(i, t), apparent power S(i, t), effective power P(i, t), effective current I(i, t), effective voltage V(i, t) and vibration frequency a(i, t); where i is the device serial number of all monitored electromechanical devices, i=1, 2, ..., n; n is the total number of all monitored electromechanical devices; and t is the time of data collection.

[0076] Additional data collection is performed based on the device type. The specific process of packaging the data collected from different devices according to a specific transmission protocol is as follows:

[0077] Get the device type of each electromechanical device i, including motor, generator, compressor, fan, and transformer.

[0078] For electromechanical devices with motors, additional status data is collected: the operating mode identifier C1(i, t) and the load status identifier C2(i, t). This data is organized in JSON format. The MQTT protocol is used for data transmission, and all collected status data for these electromechanical devices is sent to the edge data processing and optimization module. Operating modes include: running, standby, and stopped; load status identifiers include: abnormal load and normal load. The operating mode identifier C1(i, t) uses 1, 0, and -1 to represent running, standby, and stopped, respectively; the load status identifier C2(i, t) uses 0 and 1 to represent abnormal load and normal load, respectively.

[0079] For electromechanical devices with the device type of generator, additional status data is obtained: input power P_in(i, t) and output frequency P_out(i, t), and this status data is organized in JSON format. Using the AMQP protocol for data transmission, all collected status data for electromechanical devices with the device type of motor is sent to the edge data processing and optimization module.

[0080] For compressor-type electromechanical devices, additional status data is collected: pressure F(i, t), flow rate Q(i, t), and compression efficiency η(i, t). This data is organized in JSON format. MQTT is used for data transmission, and all collected status data for compressor-type electromechanical devices is sent to the edge data processing and optimization module.

[0081] For electromechanical devices of the wind turbine type, additional status data is collected: speed ω(i, t) and output power P_out(i, t). This data is organized in JSON format. HTTP is used for data transmission, and all collected status data for electromechanical devices of the wind turbine type is sent to the edge data processing and optimization module.

[0082] For electromechanical devices with transformers, additional status data is collected: input voltage Vi_in(i, t), output voltage V_out(i, t), and load factor L(i, t). This data is organized in JSON format. Using AMQP for data transmission, all collected status data for electromechanical devices with transformers is sent to the edge data processing and optimization module.

[0083] It's important to note that the MQTT protocol is suitable for scenarios requiring low latency and high reliability, and is particularly well-suited for real-time transmission of large-scale device status. The AMQP protocol supports complex message routing and high throughput, making it suitable for high-load, multi-device environments. The HTTP protocol, due to its ubiquity and ease of use, easily integrates with cloud platforms and edge computing systems, facilitating remote management and monitoring. Collecting and organizing different status data tailored to the specific needs of different device types makes monitoring data for each device more precise and relevant, improving the accuracy and efficiency of data analysis.

[0084] The edge data processing and optimization module is responsible for receiving status data from the data acquisition module, performing preliminary real-time processing of the data according to the preset control algorithm, and performing feature extraction and performance evaluation on the collected status data of each electromechanical device.

[0085] The specific process of feature extraction of the collected status data of each electromechanical device is as follows:

[0086] Extract key features from the status data of each electromechanical device, including vibration frequency spectrum features, current harmonic features, and power factor features.

[0087] The specific process of vibration frequency spectrum feature extraction is:

[0088] Input the vibration frequency a(i, t) of each electromechanical device i into the formula: The vibration frequency of each electromechanical device is converted from a time domain signal to a frequency domain signal through fast Fourier transform. Here, f is the frequency obtained after Fourier transform, X(f) is the transformed frequency domain signal, t is time, and j is the imaginary unit.

[0089] The spectrum features of the transformed frequency domain signal X(f) are extracted to obtain the fundamental frequency f0, peak frequency fmax and frequency components f1, f2, f3, ..., fmax.

[0090] Extract the energy E(f0) of the fundamental frequency f0 and the energies E(f1), E(f2), E(f3), ..., E(fmax) of the frequency components f1, f2, f3, ..., fmax. The energy calculation formula for the frequency components is:

[0091] ; Where m is the frequency component index, fm=f1,f2,f3,...,fmax;

[0092] The specific process of current harmonic feature extraction is:

[0093] Harmonic components are extracted from the effective current I(i, t) through Fourier transform, including the fundamental current I1(i) and higher harmonic currents Ik(i)=I2(i), I3(i), ..., IK(i); where k=1, 2, ..., K; K is the maximum order of the extracted harmonic current.

[0094] By preset formula Calculate the total harmonic distortion THD(i) of each electromechanical device.

[0095] The specific process of power factor feature extraction is:

[0096] Input the apparent power S(i, t) and effective power P(i, t) of each electromechanical device into the formula:

[0097] , calculate the power factor PF(i, t) of each electromechanical device i at time t.

[0098] By default formula:

[0099]

[0100] Calculate the first fluctuation coefficient of each electromechanical device i1 between the current time t and time tN and the second coefficient of fluctuation .

[0101] The first fluctuation coefficient represents the second-order moment of the power factor PF(i, t) within the past preset time, and the second fluctuation coefficient represents the maximum instantaneous rate of change of the power factor PF(i, t) within the past preset time. N is the preset power factor comparison monitoring period, and q is the timestamp index, ranging from the historical time t-N+1 to the current time t.

[0102] The central control signal matching and scheduling module is responsible for storing the first and second support vector machine models, which respectively detect abnormal conditions based on status data and extracted vibration frequency energy. Fault diagnosis and performance evaluation are performed based on the status data and extracted key features of the electromechanical devices. Signal logic matching is performed based on the calculation results of the edge data processing and optimization module to generate control signals uniquely corresponding to each electromechanical device to adjust its operating status.

[0103] The specific process of fault diagnosis and performance evaluation is as follows:

[0104] Obtain all state data of each electromechanical device i at time t to form a state data vector X1(i)={T(i, t), H(i, t), S(i, t), P(i, t), I(i, t), V(i, t), a(i, t), C1(i, t), C2(i, t), P_in(i, t), P_out(i, t), F(i, t), Q(i, t), η(i, t), ω(i, t), Vi_in(i, t), V_out(i, t), L(i, t)}; the value of the data not collected is recorded as 0.

[0105] Obtain the frequency energy E(f1), E(f2), E(f3), ..., E(fmax) of the vibration components extracted from each electromechanical device i, and form a vibration frequency energy vector X2(i) = {E(f1), E(f2), E(f3), ..., E(fmax)};

[0106] Input the state data vector X1(i) into the first support vector machine model:

[0107] ;

[0108] Obtain the first judgment label Y1(i) representing the abnormal state of the surface data; where W1 and b1 are the decision hyperplane normal vector and weight vector of the first support vector machine, respectively, which are determined by the training data;

[0109] Input the vibration frequency energy vector X2(i) into the second support vector machine:

[0110]

[0111] Obtain the second judgment label Y2(i) representing the abnormal state of the vibration frequency; where W2 and b2 are the decision hyperplane normal vector and weight vector of the second support vector machine, respectively, which are determined by the training data;

[0112] Among them, the first judgment label Y1 (i) and the second judgment label Y2 (i) are variables with a value range of 1 to 100. The larger the value, the greater the probability of an abnormality, and the smaller the value, the closer the electromechanical equipment is to normal use.

[0113] The process of training the first support vector machine model and the second support vector machine to determine the specific values ​​of W1, b1, W2 and b2 is as follows:

[0114] Perform learning data sampling, mark a set of state data vectors X1(i) and vibration frequency energy vectors X2(i) from the same electromechanical device i, and preset their corresponding standard output values and If the electromechanical equipment i fails, and are all 0; otherwise, and Both are 1.

[0115] The state data vector X1(i) and the vibration frequency energy vector X2(i) obtained by sampling the learning data are input into the first support vector machine model and the second support vector machine to obtain the output first judgment label Y1(i) and the second judgment label Y2(i).

[0116] Preset the initial values ​​of w1, w2, b1 and b2, and define the optimization objective function: , perform joint optimization operations on w1, w2, b1 and b2.

[0117] C is the preset regularization parameter that controls the degree of penalty for misclassification. and are the first and second slack variables, respectively, representing the controlled distance between the data point and the decision boundary in the first support vector machine model and the second support vector machine.

[0118] The convex optimization problem contained in the optimization objective function is solved by the Lagrange multiplier method to obtain the specific values ​​of w1, w2, b1 and b2.

[0119] The final specific values ​​of w1, w2, b1 and b2 are brought back to the first support vector machine model and the second support vector machine to complete the final training.

[0120] The specific process of signal logic matching is:

[0121] A first judgment label Y1 (i) and a second judgment label Y2 (i) of each electromechanical device i are obtained.

[0122] Obtain the total harmonic distortion THD(i) of each electromechanical device i;

[0123] Get the first fluctuation coefficient of each electromechanical device i and the second coefficient of fluctuation ;

[0124] If the first judgment tag Y1(i) of the electromechanical device i is greater than the preset threshold, and the second judgment tag Y2(i) is greater than the preset threshold, and the first fluctuation coefficient is greater than the preset threshold, and the second fluctuation coefficient is greater than the preset threshold, If the total harmonic distortion THD(i) is greater than the preset threshold, then the maintenance signal of the electromechanical equipment i is output;

[0125] If the first judgment tag Y1(i) of the electromechanical device i is greater than the preset threshold, and the second judgment tag Y2(i) is greater than the preset threshold. At the same time, if the following conditions are met: the first fluctuation coefficient is greater than the preset threshold, or the second fluctuation coefficient is greater than the preset threshold, If the value of the total harmonic distortion THD(i) is greater than the preset threshold, or the total harmonic distortion THD(i) is greater than the preset threshold, a continuous monitoring signal of the electromechanical device i is output;

[0126] If the first determination tag Y1 (i) of the electromechanical device i is identified to be less than or equal to the preset threshold, and the second determination tag Y2 (i) is less than or equal to the preset threshold, a normal operation signal of the electromechanical device i is output.

[0127] The execution and feedback module executes specific maintenance optimization operations according to the control signals uniquely corresponding to each electromechanical device, and provides real-time feedback on the operating status of the electromechanical device during the execution of the maintenance optimization operation.

[0128] For electromechanical equipment i that generates maintenance signals, immediately notify the administrator to organize personnel to carry out repairs and equipment maintenance.

[0129] For electromechanical equipment i that generates continuous monitoring signals, the administrator is immediately notified to organize personnel to conduct manual evaluation of the operating status, cleaning, and fault inspection.

[0130] For electromechanical equipment i that generates normal operating signals, no unnecessary operations are taken.

[0131] It should be understood that the terms “include” and “comprising” used in the specification and claims of the present disclosure indicate the presence of described features, integers, steps, operations, elements and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or collections thereof.

[0132] It should also be understood that the terms used in this disclosure are for the purpose of describing particular embodiments only and are not intended to limit the disclosure. As used in this disclosure and the claims, the singular forms "a," "an," and "the" are intended to include the plural forms unless the context clearly indicates otherwise. It should also be further understood that the term "and / or" used in this disclosure and the claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations;

[0133] The preferred embodiments of the present invention disclosed above are intended only to help illustrate the present invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the present invention to specific embodiments. Obviously, many modifications and variations are possible based on the contents of this specification. These embodiments are selected and described in detail in this specification to better explain the principles and practical applications of the present invention, thereby enabling those skilled in the art to better understand and utilize the present invention. The present invention is limited only by the claims and their full scope and equivalents.

Claims

1. The electromechanical equipment optimization control system based on edge computing includes a data acquisition module, an edge data processing and optimization module, and a central control signal matching and scheduling module, which is characterized by ； The data acquisition module is responsible for collecting sensor signals from electromechanical devices in real time and converting them into status data for each electromechanical device, including ambient temperature, humidity, power, current, voltage, and vibration data. It also performs additional data collection based on the device type, packages the data collected from different electromechanical devices according to a specific transmission protocol, and sends it to the edge data processing and optimization module via the communication network. The edge data processing and optimization module is responsible for receiving status data from the data acquisition module and performing preliminary real-time processing on the data according to the preset control algorithm. It also extracts features and performs performance evaluation on the status data of each electromechanical device. It extracts key features from the status data of each electromechanical device, including vibration frequency spectrum features, current harmonic features, and power factor features. The central control signal matching and scheduling module is responsible for storing the first and second support vector machine models, which respectively perform abnormal state extraction based on the state data and the extracted vibration frequency energy; and performing fault diagnosis and performance evaluation based on the state data of the electromechanical equipment and the extracted key features; Signal logic matching is carried out based on the calculation results of the edge data processing and optimization module to generate control signals uniquely corresponding to each electromechanical device to adjust the working status of each electromechanical device.

2. The electromechanical equipment optimization control system based on edge computing according to claim 1 is characterized in that: It also includes an execution and feedback module, which executes specific maintenance optimization operations according to the control signals uniquely corresponding to each electromechanical device, and provides real-time feedback on the operating status of the electromechanical device during the execution of the maintenance optimization operation.

3. The electromechanical equipment optimization control system based on edge computing according to claim 1, characterized in that: The collected status data of electromechanical equipment include ambient temperature, ambient humidity, apparent power, effective power, effective current, effective voltage and vibration frequency; Perform additional data collection based on device type, and package the data collected from different devices according to specific transmission protocols.

4. The electromechanical equipment optimization control system based on edge computing according to claim 3 is characterized in that: Additional data collection is performed based on the device type. The specific process of packaging the data collected from different devices according to a specific transmission protocol is as follows: Obtain the device type of each electromechanical device i, including motor, generator, compressor, fan, and transformer; For electromechanical devices whose device type is a motor, obtain additional status data of the device: operating mode identifier and load status identifier, and organize the status data in JSON format; Using the MQTT protocol for data transmission, all collected status data about electromechanical devices with the motor type as the device type is sent to the edge data processing and optimization module. The operating modes include: running, standby, and stopped; the load status identifiers include: abnormal load and normal load. The operating mode identifier uses 1, 0, and -1 to represent running, standby, and stopped respectively; the load status identifier uses 0 and 1 to represent abnormal load and normal load respectively. For electromechanical devices with the device type of generator, additional status data is obtained: input power and output frequency, and the status data is organized in JSON format. The AMQP protocol is used for data transmission, and all the status data collected about electromechanical devices with the device type of motor is sent to the edge data processing and optimization module. For electromechanical devices with the compressor type, additional status data is obtained: pressure, flow, and compression efficiency, and the status data is organized in JSON format. The MQTT protocol is used for data transmission, and all collected status data about the electromechanical devices with the compressor type is sent to the edge data processing and optimization module. For electromechanical devices with the device type of wind turbine, additional status data is obtained: speed and output power, and the status data is organized in JSON format. The HTTP protocol is used for data transmission, and all the collected status data about the electromechanical devices with the device type of wind turbine is sent to the edge data processing and optimization module. For electromechanical equipment with the device type of transformer, additional status data is obtained: input voltage, output voltage and load rate, and the status data is organized in JSON format; the AMQP protocol is used for data transmission, and all collected status data about the electromechanical equipment with the device type of transformer is sent to the edge data processing and optimization module.

5. The electromechanical equipment optimization control system based on edge computing according to claim 1, characterized in that: The specific process of current harmonic feature extraction and power factor feature extraction is as follows: The specific process of current harmonic feature extraction is as follows: extracting harmonic components from the effective current through Fourier transform, including fundamental current and higher harmonic current; calculating the total harmonic distortion of each electromechanical device through formula; The specific process of power factor feature extraction is as follows: the apparent power and effective power of each electromechanical device are input into the formula to calculate the power factor of each electromechanical device i at time t; Calculate the first fluctuation coefficient and the second fluctuation coefficient of each electromechanical device i1 between the current time t and the time tN; The first fluctuation coefficient represents the second-order moment of the power factor in the past preset time, and the second fluctuation coefficient represents the maximum instantaneous change rate of the power factor in the past preset time; N is the preset power factor comparison monitoring period.

6. The electromechanical equipment optimization control system based on edge computing according to claim 1, characterized in that: The specific process of fault diagnosis and performance evaluation is as follows: Obtain all state data of each electromechanical device i at time t to form a state data vector; obtain the frequency energy of the vibration component extracted from each electromechanical device i to form a vibration frequency energy vector; Inputting the state data vector into the first support vector machine model to obtain a first judgment label representing the abnormal state of the surface data; inputting the vibration frequency energy vector into the second support vector machine to obtain a second judgment label representing the abnormal state of the vibration frequency; wherein the first judgment label and the second judgment label are both variables with a value range of 1 to 100; The first support vector machine model and the second support vector machine are trained.

7. The electromechanical equipment optimization control system based on edge computing according to claim 1, characterized in that: The specific process of training the first support vector machine model and the second support vector machine is as follows: Perform learning data sampling, mark a set of state data vectors and vibration frequency energy vectors from the same electromechanical device i, and preset their corresponding standard output values ​​and; Inputting the state data vector and the vibration frequency energy vector obtained by sampling the learning data into the first support vector machine model and the second support vector machine to obtain the output first judgment label and the second judgment label; Presetting initial values ​​of the decision hyperplane normal vector and the weight vector of the first support vector machine and the second support vector machine, and defining an optimization objective function: performing a joint optimization operation on the decision hyperplane normal vector and the weight vector of the first support vector machine and the second support vector machine; The convex optimization problem contained in the optimization objective function is solved by the Lagrange multiplier method to obtain the specific values ​​of the decision hyperplane normal vector and the weight vector of the first support vector machine and the second support vector machine; The specific values ​​of the decision hyperplane normal vector and weight vector of the first support vector machine and the second support vector machine are brought back to the first support vector machine model and the second support vector machine to complete the final training.

8. The electromechanical equipment optimization control system based on edge computing according to claim 1, characterized in that: The specific process of signal logic matching is: Obtain a first judgment tag and a second judgment tag for each electromechanical device i; Obtaining the total harmonic distortion of each electromechanical device i; Obtain a first fluctuation coefficient and a second fluctuation coefficient of each electromechanical device i; If the first judgment tag of the electromechanical device i is identified to be greater than a preset threshold, the second judgment tag is greater than a preset threshold, the first fluctuation coefficient is greater than a preset threshold, the second fluctuation coefficient is greater than a preset threshold, and the total harmonic distortion is greater than a preset threshold, then a maintenance signal for the electromechanical device i is output; If the first judgment tag of the electromechanical device i is greater than the preset threshold, and the second judgment tag is greater than the preset threshold; at the same time, if the following conditions are met: the first fluctuation coefficient is greater than the preset threshold, or the second fluctuation coefficient is greater than the preset threshold, or the total harmonic loss is greater than the preset threshold, then a continuous monitoring signal of the electromechanical device i is output; If the first determination tag of the electromechanical device i is identified to be less than or equal to the preset threshold, and the second determination tag is less than or equal to the preset threshold, a normal operation signal of the electromechanical device i is output.

9. The electromechanical equipment optimization control system based on edge computing according to claim 8, characterized in that: The specific process of performing maintenance optimization operations is as follows: For electromechanical equipment i that generates maintenance signals, immediately notify the administrator to organize personnel to carry out repairs and equipment maintenance; For electromechanical equipment i that generates continuous monitoring signals, the administrator is immediately notified to organize personnel to conduct manual evaluation of the operating status, cleaning, and fault inspection.

Images