Power utilization potential safety hazard supervision method and system based on power utilization data cloud platform

Abstract

The invention relates to the technical field of power system safety monitoring, in particular to an electricity utilization potential safety hazard supervision method and system based on an electricity utilization data cloud platform. Multi-dimensional electricity utilization data are collected in real time and preprocessed, and the preprocessed data are subjected to multi-source fusion in combination with a data fusion algorithm; carrying out hidden danger detection on the fused data by utilizing a multi-layer intelligent learning model; a dynamic scoring mechanism is introduced, multi-dimensional scoring is performed on a detection result, hidden danger risk levels are divided, a multi-level early warning mechanism is triggered based on the risk levels, and automatic emergency response is implemented; potential electricity utilization safety hazards can be recognized in time, early warning is achieved, and lagged detection and processing of a traditional method are avoided. By means of intelligent hidden danger detection and dynamic risk assessment, the monitoring accuracy and efficiency are improved, human intervention is reduced, and the timeliness of emergency response is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of power system safety monitoring, and in particular to a method and system for supervising power safety hazards based on a power consumption data cloud platform. Background Art

[0002] With the rapid development of society and the advancement of industrialization, the demand for electricity is increasing, and the scale and complexity of the power system are also expanding. The operating environment of power equipment and power networks has become more and more complex, and the potential safety hazards of electricity use are becoming increasingly severe. In the fields of industry, commerce and residents' lives, any failure of power equipment or failure of the power system may lead to catastrophic consequences, such as electrical fires, equipment damage, power outages and other serious events. Therefore, how to monitor and manage power safety hazards in real time, accurately and effectively has become an important issue that needs to be solved in the power industry. Traditional power safety management methods mainly rely on manual inspections and regular testing. These methods are not only inefficient, but also have large blind spots and lags, making it difficult to respond and deal with safety hazards in a timely manner. With the development of the Internet of Things, big data, cloud computing and artificial intelligence technologies, traditional safety management methods can no longer meet the needs of modern power systems for real-time monitoring and intelligent early warning. Intelligent monitoring systems based on big data and cloud platforms can effectively improve the safety and reliability of power systems by collecting power consumption data in real time, performing data analysis and pattern recognition.

[0003] In the prior art, some intelligent power monitoring systems have begun to be applied, but most of them focus on the monitoring of a single device or a simple alarm system, lacking comprehensive analysis and dynamic risk assessment of the entire power consumption environment, and failing to achieve functions such as multi-dimensional data fusion, intelligent learning and predictive analysis. In addition, some existing early warning mechanisms lack flexibility and automatic response when dealing with sudden risks, often relying on manual intervention, and cannot achieve real-time and efficient hidden danger investigation and emergency response. In order to make up for the shortcomings of the prior art, the present invention proposes a method for supervising power safety hidden dangers based on a power consumption data cloud platform. The method collects multi-dimensional power consumption data in real time, combines data fusion algorithms to perform comprehensive processing of multi-source data, uses a multi-layer intelligent learning model to detect hidden dangers, combines a dynamic scoring mechanism to perform multi-dimensional scoring of hidden danger risks, and triggers a multi-level early warning mechanism based on risk levels to finally achieve comprehensive supervision of power safety hidden dangers. By introducing cloud platforms and artificial intelligence technologies, the present invention not only improves the accuracy and real-time performance of hidden danger monitoring, but also can automatically respond to and handle various sudden hidden dangers, providing a strong guarantee for the safe operation of the power system. Summary of the invention

[0004] In view of the above problems existing in the prior art, the present invention is proposed.

[0005] Therefore, the technical problem to be solved by the present invention is: to provide a method for supervising electricity safety hazards based on an electricity data cloud platform, which solves the lag and inefficiency problems existing in traditional electricity safety monitoring methods by integrating multi-dimensional electricity data collection, data fusion, intelligent learning, dynamic scoring and multi-level early warning mechanisms; the method aims to monitor the safety status of electrical equipment and power systems in real time and comprehensively, identify potential safety hazards in a timely manner, and predict hazard risks through intelligent analysis, thereby improving the safety and reliability of the power system; through automated hazard detection and emergency response mechanisms, the present invention can significantly reduce human intervention and response delays, optimize emergency handling efficiency, reduce the risk of accidents, and provide the power industry with an intelligent, precise and sustainable electricity safety supervision solution.

[0006] To solve the above technical problems, the present invention provides the following technical solutions: a method for supervising electricity safety hazards based on an electricity data cloud platform, comprising: real-time collection of multi-dimensional electricity data and preprocessing; using a data fusion algorithm to perform multi-source data fusion on the preprocessed data; using a multi-layer intelligent learning model to perform hazard detection on the fused data; introducing a dynamic scoring mechanism to perform multi-dimensional scoring on the hazard detection results and divide the hazard risk level; based on the hazard risk level, using a multi-level early warning mechanism for automated emergency response.

[0007] As a preferred solution of the method for supervising electricity safety hazards based on an electricity data cloud platform described in the present invention, the multi-dimensional electricity data includes voltage data, current data, power data, temperature data, humidity data, vibration data, equipment operation status data, power load data and environmental gas parameter data.

[0008] As a preferred solution of the method for supervising electricity safety hazards based on an electricity data cloud platform described in the present invention, the preprocessing includes using a local anomaly detection algorithm to screen multi-dimensional electricity data and eliminate irrelevant data and noise data.

[0009] As a preferred solution of the method for supervising electricity safety hazards based on an electricity data cloud platform described in the present invention, the data fusion algorithm is obtained by weighted summing the dimensional data; subtracting the mean of normal data from the dimensional data to obtain the data deviation; multiplying the data deviation by the corresponding dimensional data feature enhancement coefficient; dividing the product result by the product of the standard deviation of the normal data of this dimension and the global abnormal influence coefficient; summing the calculation results of all dimensions to obtain the final data fusion result.

[0010] As a preferred solution of a method for supervising potential electrical safety hazards based on an electrical consumption data cloud platform according to the present invention, wherein: the multi-layer intelligent learning model includes using the fused electrical consumption data vector as the initial input, adopting a multi-layer neural network architecture; adopting a hierarchical execution calculation process, using the fused electrical consumption data vector as the input in the first layer, using the output result of the previous layer as the input in the middle layer, and finally using an activation function to process the output value, mapping the calculation result to an interval to generate the probability of potential hazards, and setting a hazard determination threshold;

[0011] When the probability value is greater than or equal to the threshold, it is determined that there are potential safety hazards in the current electrical consumption environment. When the probability value is less than the threshold, it is determined that the current electrical consumption environment is in a safe state.

[0012] As a preferred solution of a method for supervising potential electrical safety hazards based on an electrical consumption data cloud platform according to the present invention, wherein: the dynamic scoring mechanism calculates the average value of the historical hazard probability within the calculation period by setting the monitoring period parameter, calculates the difference between the current hazard probability and the previous monitoring moment, and adopts a historical data weight value, a real-time change weight value, and a comprehensive adjustment coefficient;

[0013] When the score is lower than the first risk threshold, it is determined as a low risk, and the normal monitoring process is maintained; when the score reaches the first risk threshold but does not exceed the second risk threshold, it is determined as a medium risk, and the system automatically triggers an interface warning sign; when the score breaks through the second risk threshold, it is determined as a high risk, and the system immediately activates the emergency plan;

[0014] When the score enters the medium and high risk range from the low risk range, the system synchronously activates the audible and visual alarm device; when the score drops from the high risk range to the medium and low risk range, the system automatically turns off the alarm and generates a risk cancellation notice; during the state conversion process, the system maintains real-time data tracking and dynamically updates the risk assessment result.

[0015] As a preferred solution of a method for supervising potential electrical safety hazards based on an electrical consumption data cloud platform according to the present invention, wherein: the multi-level early warning mechanism generates an early warning response level value through a weighted algorithm based on the hazard level score, the potential risk propagation range, and the equipment importance weight;

[0016] When the early warning response level value reaches the first critical ratio, it is a low risk, and the equipment inspection process is started; when the early warning response level value breaks through the second critical ratio, it is a medium risk, and an equipment maintenance instruction is automatically sent; when the early warning response level value exceeds the third critical ratio, it is a high risk, and the full-automatic emergency response system is activated.

[0017] Another object of the present invention is to provide a power consumption safety hazard supervision system based on a power consumption data cloud platform, which realizes comprehensive monitoring and automated management of power consumption safety hazards by integrating functional modules such as data collection, data fusion, hazard detection, risk assessment, and multi-level early warning; the system can obtain and analyze multi-dimensional power consumption data in real time, use intelligent algorithms to early warn of potential hazards, dynamically evaluate the power consumption risk level, and automatically trigger emergency response measures according to the risk level, thereby greatly improving the safety and operation efficiency of the power system; the system has a high degree of automation and intelligence, can effectively reduce human operation errors, improve the emergency response speed, and provide a more scientific and efficient solution for the safety supervision of the power industry.

[0018] To solve the above technical problems, the present invention provides the following technical solutions: A power consumption safety hazard supervision system based on a power consumption data cloud platform, comprising: a data collection module, a data fusion module, a hazard detection module, a risk assessment module, and an early warning response module;

[0019] The data collection module collects multi-dimensional power consumption data in real time and performs preprocessing;

[0020] The data fusion module performs multi-source data fusion on the preprocessed data using a data fusion algorithm;

[0021] The hazard detection module uses a multi-layer intelligent learning model to detect hazards in the fused data;

[0022] The risk assessment module introduces a dynamic scoring mechanism to perform multi-dimensional scoring on the hazard detection results and divide the hazard risk levels;

[0023] The early warning response module performs automated emergency response using a multi-level early warning mechanism based on the hazard risk levels.

[0024] A computer device includes a memory and a processor, the memory stores a computer program, and is characterized in that when the processor executes the computer program, the steps of a power consumption safety hazard supervision method based on a power consumption data cloud platform as described above are implemented.

[0025] A computer-readable storage medium stores a computer program thereon, and is characterized in that when the computer program is executed by a processor, the steps of a power consumption safety hazard supervision method based on a power consumption data cloud platform as described above are implemented.

[0026] The beneficial effects of the present invention: Real-time monitoring and early warning: By collecting multi-dimensional power consumption data in real time and combining data fusion and intelligent learning algorithms, the present invention can timely identify power consumption safety hazards, realize early warning of potential risks, and avoid the lagged detection and treatment of hazards in traditional methods.

[0027] Intelligent Hidden Danger Detection: By using a multi-layer intelligent learning model to detect hidden dangers in the fused data, it can automatically identify complex power consumption equipment failures or safety hazards, improving the accuracy and efficiency of detection.

[0028] Dynamic Risk Assessment and Automatic Response: Introduce a dynamic scoring mechanism to score the hidden danger risks in multiple dimensions, and automatically trigger different levels of emergency responses according to the risk levels, reducing the error rate and response delay of human intervention, and improving the timeliness and accuracy of emergency responses.

[0029] Improve the Safety and Reliability of the Power System: By comprehensively evaluating the hidden danger risks and taking real-time monitoring and emergency measures, the present invention can effectively reduce the occurrence of power accidents, ensure power consumption safety, and improve the stability and reliability of the power system.

[0030] Reduce Operating Costs: The system has a high degree of automation and intelligence, reducing the dependence on manual inspections, thereby reducing the maintenance costs of power equipment operation and improving management efficiency. Brief Description of the Drawings

[0031] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0032] Figure 1 It is a flowchart of a method for supervising power consumption safety hidden dangers based on a power consumption data cloud platform provided by an embodiment of the present invention.

[0033] Figure 2 It is a schematic diagram of hidden danger risk division of a method for supervising power consumption safety hidden dangers based on a power consumption data cloud platform provided by an embodiment of the present invention.

[0034] Figure 3 It is a system module diagram of a method for supervising power consumption safety hidden dangers based on a power consumption data cloud platform provided by an embodiment of the present invention.

[0035] Figure 4 It is a visual comparison diagram of a method for supervising power consumption safety hidden dangers based on a power consumption data cloud platform provided by an embodiment of the present invention. Detailed Embodiments

[0036] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0037] Example 1, reference Figure 1 - Figure 2 , is an embodiment of the present invention, which provides a method for monitoring power safety hazards based on a power consumption data cloud platform, including:

[0038] S1: Real-time collection of multi-dimensional electricity consumption data and pre-processing.

[0039] It should be noted that if Figure 1 As shown in S1, the multi-dimensional electricity consumption data includes voltage data, current data, power data, temperature data, humidity data, vibration data, equipment operation status data, power load data and environmental gas parameter data.

[0040] Furthermore, preprocessing includes using local anomaly detection algorithms to filter multi-dimensional electricity consumption data and remove irrelevant and noisy data;

[0041] Among them, the local anomaly detection algorithm is expressed as:

[0042]

[0043] Among them, S(x) represents the anomaly scoring function, that is, the degree of anomaly of the input data, N represents the number of dimensions of the data, and x i represents the value of the input multi-dimensional electricity consumption data in the i-th dimension, μ represents the mean of normal data, σ represents the standard deviation of normal data, and w i Represents the weight of the i-th dimension data, which is used to reflect the importance of different dimensions in anomaly detection. max(0,|x i -μ|-σ) is used to filter out changes less than one standard deviation and only calculate scores for abnormal data with large deviations.

[0044] Furthermore, the local anomaly detection algorithm evaluates the data of each dimension based on the anomaly scoring function. If the data of a certain dimension deviates from the normal range, it is marked as potential abnormal data. The data with a small deviation and no significant impact are filtered out, and only the abnormal data with a large deviation from the normal range is retained, thereby improving data processing efficiency and reducing false alarm rate, ensuring the dominant role of key data dimensions (such as voltage, current, etc.) in anomaly detection.

[0045] S2: Use data fusion algorithm to fuse multi-source data after preprocessing.

[0046] It should be noted that, as shown in S2 Figure 1 , the data fusion algorithm is expressed as:

[0047]

[0048] Among them, F(x) represents the data fusion result, and μ i represents the mean of the i-th dimensional normal data, and σ i represents the standard deviation of the i-th dimensional normal data, α i represents the data feature enhancement coefficient of the i-th dimension, and β represents the global anomaly influence coefficient, which is used to adjust the comprehensive influence of the full-dimensional anomaly distribution on the fusion result. represents normalizing the i-th dimensional data to eliminate the influence of dimension. represents the overall deviation of all dimensional data, reflecting the anomaly degree of the overall data distribution.

[0049] Furthermore, the data fusion result F(x) calculates the data of each dimension through weighting, comprehensively considering the importance of each dimension in anomaly detection; the data of each dimension will refer to the mean μ i of the normal data of that dimension and the standard deviation σ i when calculating to ensure that the data of different dimensions maintain a consistent dimension standard during fusion; the feature enhancement coefficient α i is used to amplify the dimensions that have a greater impact on potential hazard detection, further improving the sensitivity of the fusion result, and ensuring that the contributions of key dimensions (such as current and voltage) to risk assessment are sufficiently reflected.

[0050] Even further, the global anomaly influence coefficient β is a global adjustment factor, which is used to correct the influence of the abnormal distribution of the overall data on the fusion result; it helps to eliminate the deviation brought by the excessive or too small anomaly of individual dimensions to the fusion result, making the fusion result more balanced.

[0051] S3: Use a multi-layer intelligent learning model to detect potential hazards in the fused data.

[0052] It should be noted that, as shown in S3 Figure 1 , the multi-layer intelligent learning model is expressed as:

[0053]

[0054] Among them, D(x) represents the probability of potential hazard existence, x R represents the fused power consumption data vector, which is the comprehensive power consumption data vector obtained by the data fusion algorithm from the multi-dimensional power consumption data after fusion preprocessing, L represents the number of layers in the model, and γ l represents the output scaling factor of the l-th layer, and W lDenote the weight matrix of the l-th layer, b l Denote the bias vector of the l-th layer, ReLU denotes the activation function, and σ denotes the Sigoid activation function, which is used to convert the output into a probability format.

[0055] Furthermore, ReLU is defined as ReLU(z) = max(0, z), and for the l-th layer, calculate z l = W l ·x R + b l , which is used as the input data in the first layer and the output of the previous layer in subsequent layers, and the final layer outputs z L After being processed by the Sigoid activation function σ, the probability of potential hazards is obtained; set the hazard judgment threshold to 0.5. If D(x) ≥ 0.5, it is judged that there are potential hazards in the current environment, and corresponding warnings are triggered and the processing process is entered; if D(x) < 0.5, it is judged that there are no hazards in the current environment, and monitoring can be maintained.

[0056] Furthermore, each layer l of the model processes the input data through the weight matrix W l and the bias vector b l , and after the linear transformation W l ·x R + b l , then a non-linear mapping is performed through the activation function ReLU to enhance the model's fitting ability for complex patterns; and the output of each layer is adjusted by the scaling factor γ l to optimize the performance of the model in each layer; finally, in the last layer, the output value z l of the model will be converted into a probability value through the Sigmoid activation function σ, indicating the possibility of potential hazards in the current input data; in order to improve the generalization ability of the model and its ability to handle different types of hazards, backpropagation and parameter updates will be performed during the training process, and the weights and biases will be continuously adjusted using historical data and labeled hazard instances; through repeated training, the model learns to distinguish between normal states and potential hazards and can identify possible danger signals in real-time data streams.

[0057] S4: Introduce a dynamic scoring mechanism to perform multi-dimensional scoring on the hazard detection results and divide the hazard risk levels.

[0058] It should be noted that, as shown in S4 Figure 1 above, the dynamic scoring mechanism is expressed as:

[0059]

[0060] Among them, R(t) represents the hazard risk level score at the current time t, and T represents the time window length. Denote the average value of historical hidden danger detection probability, ΔD denote the change rate considered for the current hidden danger, and ω1, ω2, ω3 denote weight coefficients;

[0061] Among them, ΔD is defined as ΔD = D(x) - D(x prev ), where D(x prev ) represents the hidden danger detection probability at the previous moment.

[0062] Furthermore, when 0 ≤ R(t) < 0.4, it is classified as low risk, and no additional action is required. Keep regular monitoring. By setting a lower threshold value (0.4), it can effectively distinguish between regular fluctuations and risks that really need attention, and avoid over-intervening in some normal fluctuations or minor changes; when 0.4 ≤ R(t) < 0.7, it is classified as medium risk, and the system issues a prompt message, highlighting the risk area in red on the monitoring interface. By setting the range of (0.4 - 0.7), it can ensure that in the case of medium risk, the staff has enough time and warnings to conduct inspections and prevention, and can avoid the system triggering a high-intensity emergency response when there are minor abnormalities, so as to reasonably allocate the pressure and importance of the response; when 0.7 ≤ R(t) < 1.0, it is classified as high risk, and the system triggers an alarm and executes emergency measures; if the hidden danger risk level score rises from low risk to medium risk and high risk, the system immediately issues an alarm; if the hidden danger risk level score drops from high risk to medium risk and low risk, the system terminates the alarm and prompts that the risk is lifted. By setting a relatively high threshold (0.7), it can ensure that a timely response can be made when the hidden danger is serious, avoid late response or operation delay, and also ensure that the system can be protected in time.

[0063] Even further, as Figure 2 shown, when the hidden danger risk level score is low risk, it means that there are no obvious safety hidden dangers in the current power consumption environment, and the system will maintain the regular monitoring state without taking additional actions; at this time, the system will continue to track and record power consumption data in real time to ensure that any potential problems can be discovered in time; when the hidden danger risk level score rises to medium risk, the system will issue a prompt message according to the detection results to remind the staff to pay attention to potential safety hidden dangers; in the monitoring interface, the relevant area will be highlighted in red to warn the user of the possibility of risks; at this time, the system continuously monitors the data and guides the maintenance personnel to go to the possible risk area for on-site inspection; and in the medium risk state, the system will also automatically record possible abnormal data and monitoring results for subsequent analysis; when the hidden danger risk level score reaches high risk, it means that there are serious potential hidden dangers, and the system will immediately trigger an alarm and start the preset emergency response measures; the system will send an emergency notice to relevant personnel and provide real-time data feedback to ensure that the emergency response measures can be quickly executed;

[0064] If the hidden danger risk level score rises from low risk to medium risk or high risk, the system will immediately issue an alarm to ensure that all relevant personnel can obtain early warning information in time and take necessary preventive measures to avoid accidents; and the alarm system will cover users at all levels, including managers, operation and maintenance personnel and other relevant responsible persons, to ensure that each link can respond in time; when the hidden danger risk level score drops from high risk to medium risk or low risk, the system will terminate the alarm in time according to the situation, lift the emergency state, and prompt the risk to be lifted; at this time, the system will record the changes in risk levels in detail and automatically update the risk assessment report for subsequent analysis and optimization.

[0065] S5: Based on the risk level of hidden dangers, use a multi-level early warning mechanism for automated emergency response.

[0066] It should be noted that if Figure 1 As shown in S5, the multi-level early warning mechanism is expressed as:

[0067] P(t)=κ1·H(t)+κ2·W(t)+κ3·E(t)

[0068] Among them, P(t) represents the warning response level value, H(t) represents the weighted value corresponding to the current hidden danger level score R(t), W(t) represents the potential risk propagation range, E(t) represents the equipment importance weight, κ1, κ2, κ3 represent weight coefficients, which respectively represent the impact of hidden danger risk level, propagation range and equipment importance on the warning level;

[0069] Where H(t) is defined as:

[0070]

[0071] W(t) is defined as:

[0072] W(t) = range impact coefficient × number of associated devices

[0073] E(t) is defined as:

[0074] E(t) = Important equipment impact score.

[0075] Further, when 0.7 ≤ R(t) < 1.0, it is in a high-risk state, and the system triggers an alarm. The warning level is divided according to the warning response level value. When P(t) < 0.6, a first-level warning is triggered to prompt for inspection. The threshold of 0.6 is set to avoid overreaction, reduce waste of system resources, and ensure that warnings are issued only when actual intervention is needed; when 0.6 ≤ P(t) < 0.8, a second-level warning is triggered to send maintenance information. The threshold of 0.6 defines the boundary between low risk and medium risk, while the threshold of 0.8 marks the entry into a more urgent state. The setting of these two values enables the system to balance the timeliness and accuracy of alarm triggering, and enables relevant staff to intervene according to the actual situation to avoid unnecessary warnings; when P(t) ≥ 0.8, a third-level warning is triggered to immediately trigger an automated warning response. The threshold of 0.8 is set as the boundary value for a higher risk. The purpose of this setting is to ensure that timely and sufficient responses can be obtained when the danger gradually intensifies. If the risk is too high and not controlled in time, it may cause irreparable losses to the equipment or the environment. The threshold of 0.8 is set based on the tolerance of the power system and can provide time and space to take emergency measures;

[0076] When a third-level warning is triggered, continuously monitor R(t) and check the trend of ΔR. If it continues to rise, push warning information, including the hidden danger level score R(t), the warning response level value P(t), the potential risk propagation range W(t), and the equipment importance weight E(t);

[0077] According to the equipment importance weight E(t), take emergency measures, including power-off operation, load reallocation, and equipment isolation; and record the execution status of the emergency measures taken, continuously monitor the change of R(t). If it is detected that R(t) < 0.7, the third-level warning is automatically lifted, and the maintenance information is retained and the inspection is maintained. The threshold of 0.7 is used as the threshold for approaching a high-risk warning and at the same time as the critical value for lifting the highest warning. When R(t) < 0.7, it indicates that the hidden danger risk level is also decreasing from a high risk level, meaning that the system has returned to a relatively safe state and there is no need to trigger a warning state, avoiding unnecessary emergency responses and resource waste; if it is detected that R(t) > 0.9, increase the intensity of the emergency response and activate emergency equipment. The threshold of 0.9 is selected to ensure that when approaching the limit risk, the system can take the maximum-intensity emergency measures to ensure that in the stage when the system faces the greatest potential hidden danger, the timely activation of standby facilities can effectively reduce the probability of system downtime and the occurrence of serious accidents;

[0078] Among them, the power-off operation: cut off the power supply of the high-risk area through the control module to prevent the further spread of hidden dangers; load reallocation: reallocate the loads in other areas to ensure the stability of the power supply system; equipment isolation: automatically close the control channels associated with high-risk equipment;

[0079] Where ΔR is defined as:

[0080] ΔR=R(t)-R(t-1)

[0081] Among them, ΔR represents the changing trend of the hidden danger level score, and R(t-1) represents the hidden danger level score at the previous moment.

[0082] Furthermore, when a level 1 warning is triggered, the system will prompt for routine inspections and advise staff to conduct detailed inspections of areas where problems may occur to prevent the further expansion of hidden dangers; when a level 2 warning is triggered, the system will send maintenance information to remind relevant personnel to pay attention to hidden dangers and take preventive measures, and may require preliminary inspections and repairs of equipment; when a level 3 warning is triggered, automated emergency response measures will be initiated to quickly control and eliminate hidden dangers to avoid further expansion of risks; and during the level 3 warning response process, the system will continue to monitor the trend of hidden danger changes. If the hidden danger level score continues to rise, the system will continue to push warning information, providing detailed information such as hidden danger level score, warning response level value, potential risk transmission range, and equipment importance weight, to ensure that relevant personnel obtain complete emergency information and take timely measures;

[0083] According to the equipment importance weight E(t), the system will give priority to taking emergency measures for key equipment; for example, if important equipment is threatened, the system will give priority to power-off operations, isolate the risk source, or redistribute the load to maintain the stability of the power system; the system will record the execution status of each emergency measure, and continuously track the equipment status and changes in hidden dangers. Once the hidden danger is controlled or the risk is reduced, the system will automatically lift the third-level warning and prompt that the risk has been lifted; at this time, the system continues to retain maintenance information and maintain inspections to ensure that the hidden danger is effectively dealt with; if it is monitored that the hidden danger has been re-escalated or there is a new potential risk, the system will increase the emergency response and enable backup emergency equipment (such as backup power supply, emergency cooling system) to strengthen safety protection.

[0084] The above is a schematic scheme of a method for supervising power safety hazards based on a power data cloud platform in this embodiment. It should be noted that the technical scheme of the system for supervising power safety hazards based on a power data cloud platform and the technical scheme of the above-mentioned method for supervising power safety hazards based on a power data cloud platform belong to the same concept. The details of the technical scheme of the power safety hazard supervision system based on a power data cloud platform in this embodiment that are not described in detail can all be referred to the description of the technical scheme of the above-mentioned method for supervising power safety hazards based on a power data cloud platform.

[0085] Example 2, reference Figure 3, in this embodiment, a power consumption safety hazard supervision system based on a power consumption data cloud platform is provided, which is characterized in that it includes: a data collection module, a data fusion module, a hazard detection module, a risk assessment module, and an early warning response module;

[0086] The data collection module collects multi-dimensional power consumption data in real time and performs preprocessing;

[0087] The data fusion module performs multi-source data fusion on the preprocessed data by using a data fusion algorithm;

[0088] The hazard detection module uses a multi-layer intelligent learning model to detect hazards in the fused data;

[0089] The risk assessment module introduces a dynamic scoring mechanism to perform multi-dimensional scoring on the hazard detection results and divide the hazard risk levels;

[0090] The early warning response module performs automated emergency responses based on the hazard risk levels by using a multi-level early warning mechanism.

[0091] This embodiment also provides a computing device, which is applicable to a situation of a power consumption safety hazard supervision method based on a power consumption data cloud platform, including:

[0092] A memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement a power consumption safety hazard supervision method based on a power consumption data cloud platform as proposed in the above embodiment.

[0093] This embodiment also provides a storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements a power consumption safety hazard supervision method based on a power consumption data cloud platform as proposed in the above embodiment.

[0094] The storage medium proposed in this embodiment and a power consumption safety hazard supervision method based on a power consumption data cloud platform proposed in the above embodiment belong to the same inventive concept. Technical details not described in detail in this embodiment can be referred to in the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0095] When the above-mentioned functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a 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 causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM, Read-Only Memory), random access memories (RAM, Random Access Memory), magnetic disks, or optical discs that can store program codes.

[0096] The logic and / or steps described in other ways herein, for example, can be considered as a definite sequence list of executable instructions for implementing logical functions, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch instructions from the instruction execution system, apparatus, or device and execute the instructions), or in combination with these instruction execution systems, apparatuses, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that can contain, store, communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.

[0097] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following well-known technologies in the art can be used: discrete logic circuits having logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits having appropriate combinational logic gate circuits, programmable gate arrays (PGA), field programmable gate arrays (FPGA), etc.

[0098] Example 3, refer to Figure 4 One embodiment of the present invention provides a method for supervising potential electrical safety hazards based on an electrical consumption data cloud platform. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0099] Experimental steps: Set an experimental group (a method for supervising potential electrical safety hazards based on an electrical consumption data cloud platform) and a control group (a traditional method for monitoring potential electrical safety hazards).

[0100] Collect multi-dimensional power consumption data in real time, including voltage, current, power, temperature, humidity, vibration, equipment operation status, load data, and environmental gas parameters, etc.; preprocess the collected data, remove noise and irrelevant data, and use local anomaly detection algorithms for screening.

[0101] The experimental group uses data fusion algorithms to fuse the processed multi-source data. This includes standardizing and enhancing the features of each dimension of data to optimize the hidden danger detection results; the control group does not use data fusion and processes each dimension of data separately.

[0102] The experimental group uses a multi-layer intelligent learning model to detect hidden dangers in the fused data and obtain the probability of the existence of hidden dangers; the control group uses traditional methods (such as rule-based threshold judgment) for hidden danger identification.

[0103] The experimental group conducts multi-dimensional scoring on the hidden danger risks through a dynamic scoring mechanism, updates the risk level in real time, and triggers multi-level warnings based on the risk level; the control group only issues warnings based on a pre-set fixed threshold.

[0104] Based on the risk level, the experimental group automatically triggers emergency responses (such as power off, load redistribution, equipment isolation, etc.) and updates the emergency response status in real time; the control group relies on manual intervention to execute emergency responses, and the response time is longer.

[0105] According to the experimental data (as shown in Table 1 below) and Figure 4 Analysis: Hidden danger scoring: The hidden danger scores of the experimental group (average 0.61) and the control group (average 0.71) are relatively close, but the experimental group can more accurately predict potential hidden dangers through multi-dimensional data fusion and intelligent learning models.

[0106] Response time: The emergency response time of the experimental group is significantly lower than that of the control group. The average response time of the experimental group is 3.00 seconds, while that of the control group is 5.75 seconds. The experimental group uses an automated emergency response mechanism, reducing the manual intervention time and improving the response efficiency.

[0107] False alarm rate: The false alarm rate of the experimental group is relatively low (average 0.05), while that of the control group is relatively high (average 0.13). This indicates that the method based on the cloud platform and intelligent algorithms has more advantages in reducing false alarms.

[0108] Through the combination of real-time data collection, multi-dimensional data fusion, and intelligent learning models, the experimental group can detect potential hidden dangers more accurately compared to traditional methods, reducing the false alarm rate. The emergency response time of the experimental group is significantly shorter than that of the control group, thanks to the automated warning and emergency response mechanisms. The experimental group shows a high level of automation, capable of processing data, identifying hidden dangers, and automatically responding in real time, significantly reducing the need for manual intervention.

[0109] Table 1 Experimental Data

[0110]

[0111] As Figure 4 shown: Abscissa (dimension): Each axis of the graph represents a different evaluation dimension. For example: Hidden danger score, response time, false alarm rate, equipment importance, risk level, etc. These dimensions are set according to the experimental steps of the experimental group and the control group.

[0112] Experimental group and control group: In the graph, the performance of the experimental group and the control group are represented by blue (experimental group) and purple (control group) respectively. The values of each dimension (such as hidden danger score, response time, etc.) are marked by data points on the corresponding dimension, and these points are connected by lines to form a closed graphic area.

[0113] Graphic interpretation: The size, shape and position of the graphic areas of the experimental group and the control group reflect their performance in each dimension. The larger part of the graphic area indicates that a certain group performs better in that dimension. On the contrary, the smaller area indicates a poorer performance in that dimension.

[0114] Hidden danger score: The difference in the hidden danger scores between the experimental group and the control group is relatively small, but the experimental group has a more balanced improvement in this dimension, especially in the accuracy of hidden danger identification and assessment, thanks to the use of multi-dimensional data fusion and intelligent learning models.

[0115] Response time: The response time of the experimental group is significantly lower than that of the control group. The average response time of the experimental group is 3.00 seconds, while that of the control group is 5.75 seconds, indicating that the automated early warning and emergency response mechanism of the experimental group has significantly improved the response speed and reduced the manual intervention time.

[0116] False alarm rate: The false alarm rate of the experimental group is relatively low (average 0.05), while that of the control group is relatively high (average 0.13). This means that the experimental group can more accurately judge potential hidden dangers through intelligent algorithms and real-time data collection, reducing false alarm phenomena.

[0117] Advantages of the experimental group: Automation and precision: By real-time collecting multi-dimensional data, data fusion and intelligent learning models, the experimental group not only improves the accuracy of hidden danger detection, but also updates the risk level in real time through a dynamic scoring mechanism, ensuring a fast response. Low false alarm rate: Due to the use of more advanced data processing and analysis methods, the false alarm rate of the experimental group is relatively low, which can effectively reduce the untimely system response caused by human judgment errors. Fast response: The automated emergency response mechanism enables the experimental group to trigger an emergency response within 3.00 seconds, while the control group requires a longer time for manual intervention.

[0118] Disadvantages of the control group: Dependence on manual intervention: The control group uses traditional rule-based threshold judgment and relies on manual intervention to execute emergency responses, resulting in a longer response time and a relatively high false alarm rate. Its data processing ability and emergency response mechanism lack real-time performance and accuracy. Single data processing: It does not adopt data fusion algorithms and intelligent learning models, resulting in a relatively simple judgment of potential hidden dangers and ignoring the relationships between multi-dimensional data, which limits its prediction ability.

[0119] Comprehensive analysis: The advantages of the experimental group are mainly reflected in two aspects: real-time performance and accuracy. The experimental group uses the cloud platform for data collection and processing, and optimizes hidden danger detection through an intelligent learning model. It can not only accurately identify hidden dangers but also quickly respond when potential risks occur, reducing the time of manual intervention and the false alarm rate. The control group is relatively lagging behind. Although its hidden danger score is similar to that of the experimental group, it performs poorly in key indicators such as response speed and false alarm rate. The control group relies on manual intervention and fixed thresholds for early warning, with low efficiency and failure to fully utilize the advantages of data analysis. The experimental group is significantly superior to the control group, with obvious advantages in hidden danger detection, response time, and false alarm rate. Based on real-time data processing on the cloud platform, multi-dimensional data fusion, and intelligent learning algorithms, the experimental group can quickly and accurately identify potential hidden dangers and conduct emergency responses, improving the automation level and overall safety of the system.

[0120] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered by the scope of the claims of the present invention.

Claims

1. A method for monitoring hidden dangers of power safety based on a power data cloud platform, characterized by: include: Collect multi-dimensional electricity consumption data in real time and perform pre-processing; Use data fusion algorithm to fuse multi-source data after preprocessing; Use multi-layer intelligent learning models to detect hidden dangers in fused data; Introduce a dynamic scoring mechanism to conduct multi-dimensional scoring of hidden danger detection results and classify hidden danger risk levels; Based on the risk level of hidden dangers, a multi-level early warning mechanism is used for automated emergency response.

2. A method for monitoring hidden dangers of electricity safety based on an electricity data cloud platform as claimed in claim 1, characterized in that: The multi-dimensional electricity consumption data includes voltage data, current data, power data, temperature data, humidity data, vibration data, equipment operation status data, power load data and environmental gas parameter data.

3. A method for monitoring hidden dangers of electricity safety based on an electricity data cloud platform as claimed in claim 2, characterized in that: The preprocessing includes using a local anomaly detection algorithm to screen the multi-dimensional electricity consumption data and eliminate irrelevant data and noise data.

4. A method for monitoring hidden dangers of electricity safety based on an electricity data cloud platform as claimed in claim 3, characterized in that: The data fusion algorithm is obtained by weighted summing the dimensional data; subtracting the normal data mean from the dimensional data to obtain the data deviation; multiplying the data deviation by the corresponding dimensional data feature enhancement coefficient; Divide the product result by the product of the normal data standard deviation of this dimension and the global abnormal influence coefficient; sum the calculation results of all dimensions to obtain the final data fusion result.

5. A method for monitoring hidden dangers of electricity safety based on an electricity data cloud platform as claimed in claim 4, characterized in that: The multi-layer intelligent learning model includes taking the power consumption data vector after data fusion as the initial input and adopting a multi-layer neural network architecture; The calculation process is executed in layers. The first layer takes the fused power consumption data vector as input, the middle layer takes the output result of the previous layer as input, and the last layer uses the activation function to process the output value, maps the calculation result to the interval, generates the probability of hidden danger existence, and sets the hidden danger judgment threshold. When the probability value is greater than or equal to the threshold, it is determined that the current power usage environment has safety hazards. When the probability value is less than the threshold, it is determined that the current power usage environment is in a safe state.

6. A method for monitoring hidden dangers of electricity safety based on an electricity data cloud platform as claimed in claim 5, characterized in that: The dynamic scoring mechanism sets monitoring cycle parameters, calculates the average value of historical hidden danger probabilities within the cycle, calculates the difference between the current hidden danger probability and the previous monitoring moment, and adopts historical data weight values, real-time change weight values ​​and comprehensive adjustment coefficients; When the score is lower than the first risk threshold, it is judged as low risk and the normal monitoring process is maintained; When the score reaches the first risk threshold but does not exceed the second risk threshold, it is judged as medium risk and the system automatically triggers the interface warning mark; When the score exceeds the second risk threshold, it is judged as high risk and the system immediately activates the emergency plan; When the score changes from the low-risk range to the medium-high risk range, the system will simultaneously activate the sound and light alarm device; when the score drops from the high-risk range to the medium-low risk range, the system will automatically turn off the alarm and generate a risk relief notice; During the state transition process, the system maintains real-time data tracking and dynamically updates risk assessment results.

7. A method for monitoring hidden dangers of electricity safety based on an electricity data cloud platform as claimed in claim 6, characterized in that: The multi-level warning mechanism generates a warning response level value through a weighted algorithm based on the hidden danger level score, potential risk propagation range and equipment importance weight; When the warning response level value reaches the first critical ratio, it is low risk and the equipment inspection process is started; When the early warning response level value exceeds the second critical ratio, it is a medium risk and equipment maintenance instructions are automatically sent; when the early warning response level value exceeds the third critical ratio, it is a high risk and the fully automatic emergency response system is activated.

8. A system for monitoring power safety hazards based on a power data cloud platform according to any one of claims 1 to 7, characterized in that: include: Data acquisition module, data fusion module, hidden danger detection module, risk assessment module and early warning response module; The data acquisition module collects multi-dimensional electricity consumption data in real time and performs pre-processing; The data fusion module uses a data fusion algorithm to perform multi-source data fusion on the pre-processed data; The hidden danger detection module uses a multi-layer intelligent learning model to perform hidden danger detection on the fused data; The risk assessment module introduces a dynamic scoring mechanism to perform multi-dimensional scoring on hidden danger detection results and classify hidden danger risk levels; The early warning response module uses a multi-level early warning mechanism to perform automated emergency response based on the hidden danger risk level.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of a method for supervising electricity safety hazards based on an electricity data cloud platform as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, the steps of a method for supervising electricity safety hazards based on an electricity data cloud platform as described in any one of claims 1 to 7 are implemented.

Images