Safety control method of smart power safety control cabinet based on IoT sensor

Abstract

This invention discloses a safety control method for a smart power safety control cabinet based on IoT sensors, aiming to provide comprehensive power safety monitoring and management. The method includes the following steps: Real-time monitoring of various parameters of the power system and environment is achieved by integrating current, voltage, temperature, humidity, and smoke sensors. A data processing device preprocesses and reduces noise in the collected data, and uses a topology data analysis unit and a Gaussian process regression unit for anomaly detection. The control device triggers audible and visual alarms based on the detection results and transmits the alarm information to a remote monitoring platform and mobile devices via quantum communication and a 5G network. Users can view system status and historical records through the remote platform and perform remote management and operation. This invention introduces topology data analysis and a Gaussian process regression model, combined with quantum computing and communication technologies, to improve the accuracy of anomaly detection, the reliability of data transmission, and the system's response speed, thus ensuring power safety.

Description

Technical Field

[0001] The present invention relates to the technical field of power system safety management, and in particular to a safety control method of a smart power safety control cabinet based on an Internet of Things sensor. Background Art

[0002] With the development of society and the advancement of science and technology, electricity has become an indispensable energy source in people's daily life and industrial production. However, the safety of the power system has always been an important factor affecting social stability and economic development. Especially in different application scenarios such as industry, office buildings, public facilities and homes, the failure and safety hazards of the power system may bring serious consequences, such as equipment damage, fire and even casualties. Therefore, how to effectively monitor and manage the safety of the power system has become a hot topic and difficulty in current research.

[0003] Traditional power system safety management usually relies on manual inspections and simple power protection equipment. These methods have many shortcomings. First, manual inspections are inefficient and cannot achieve real-time monitoring of the power system, which easily leads to potential problems that cannot be discovered and handled in a timely manner. Secondly, traditional power protection equipment mostly uses simple overload and short-circuit protection, which cannot comprehensively consider the impact of environmental factors on power system safety. In addition, traditional equipment usually lacks intelligent analysis and remote monitoring capabilities, making it difficult to meet the needs of modern management.

[0004] Although some existing intelligent power monitoring systems have improved monitoring efficiency to a certain extent, they still have some technical deficiencies. For example, many systems can only monitor a single or a small number of power parameters and cannot fully reflect the operating status of the power system. At the same time, most of these systems use traditional data processing and analysis methods, with low detection accuracy and reliability, making it difficult to effectively identify and warn of abnormal situations. In addition, the alarm and data transmission methods of existing systems are relatively simple, lacking efficient and reliable remote monitoring and management methods.

[0005] Therefore, how to provide a safety control method for a smart power safety control cabinet based on IoT sensors is an issue that technical personnel in this field urgently need to solve. Summary of the invention

[0006] One purpose of the present invention is to propose a safety control method for a smart electricity safety control cabinet based on Internet of Things sensors. The present invention makes full use of Internet of Things technology, sensing technology, wireless communication technology, data analysis technology, quantum computing and 5G communication technology, and describes in detail an intelligent method for real-time monitoring of power systems and environmental parameters, abnormality detection, triggering alarms, and remote monitoring and management. It has the advantages of high safety, high detection accuracy, reliable data transmission and fast system response speed.

[0007] According to an embodiment of the present invention, a safety control method for a smart power safety control cabinet based on an IoT sensor includes the following steps:

[0008] S1. Install a multi-sensor integrated device in the power system, including current sensors, voltage sensors, temperature sensors, humidity sensors and smoke sensors, to monitor various parameters of the power system and environment in real time;

[0009] S2, collects current, voltage, temperature, humidity and smoke data in real time through a multi-sensor integrated device, and transmits the data to a built-in control unit for intelligent analysis;

[0010] S3, built-in control unit uses improved Neural Ordinary Differential Equation algorithm to perform continuous time modeling and dynamic change analysis on sensor data;

[0011] S4. The improved Neural Ordinary Differential Equation algorithm is combined with the deep Q-learning algorithm to automatically optimize and adjust the security strategy according to historical data and real-time environmental changes, and realize intelligent and adaptive security control, including power outage, alarm and adjustment of power load distribution;

[0012] S5. Use quantum computing to assist anomaly detection, process and analyze large-scale sensor data in real time through the parallel computing power of quantum computing, and transmit data in interference environments;

[0013] S6. Use quantum communication and 5G networks to achieve real-time transmission of monitoring data. Users can view system status, monitoring data and alarm information through remote monitoring platforms and mobile devices, and perform remote operations and management.

[0014] S7. When an abnormal situation is detected, the system will issue an audible and visual alarm, and send the alarm information to the remote monitoring platform and the mobile devices of relevant personnel through quantum communication and 5G network;

[0015] S8. The system automatically records all monitoring data and alarm events, forms historical records, and provides data analysis functions to help users trace faults and optimize power management strategies.

[0016] Optionally, the step S3 specifically includes:

[0017] S31, collecting sensor data X(t), where X(t) includes current I(t), voltage V(t), temperature T(t), humidity H(t) and smoke concentration S(t);

[0018] S32, inputting the collected sensor data X(t) into the improved neural ordinary differential equation model:

[0019]

[0020] Where f(X(t),θ) represents the basic neural ordinary differential equation part, α is the weight coefficient used to adjust the influence of the time-dependent compensation function g(t,φ), and θ and φ are model parameters;

[0021] S33, training the improved neural ordinary differential equation model through a multi-stage optimization algorithm to determine model parameters θ and φ;

[0022] S34, using the optimized model parameters θ and φ to analyze the real-time sensor data:

[0023]

[0024] in, Represents the predicted state at time t+Δt, and calculates the dynamic changes of data through integration;

[0025] S35. Combine historical data H(X) and real-time data X(t) for comprehensive analysis to predict potential abnormal situations:

[0026]

[0027] Where P(A|X,H) is the predicted probability of anomaly, σ is the activation function, and W and b are model parameters;

[0028] S36. Based on the prediction results, the optimized model parameters θ and φ are used to continuously monitor the sensor data and analyze the dynamic changes, and the safety control strategy is updated.

[0029] Optionally, the step S4 specifically includes:

[0030] S41. Define the state s of deep Q learning t 、Action a t and the reward function r t ;

[0031] S42, define the state s t and action a t And the corresponding reward r t Input into the improved Neural ODE model;

[0032] S43, combined with the deep Q learning algorithm, by using the improved neural ordinary differential equation model to calculate the state s t and instant rewards t , the Q network parameters θ are optimized by minimizing the following loss function:

[0033]

[0034] Among them, Q(s t ,a t; θ) represents the state s t Next, perform action a t Q value, θ - is the parameter of the target Q network, γ is the discount factor, and D is the experience replay pool used to store state transition samples;

[0035] S44, apply the trained control strategy in the built-in control unit, monitor the sensor data X(t) in real time and t Choose the best action a t :

[0036]

[0037] Determine the current state s by monitoring the real-time sensor data X(t) t , and select the optimal action a based on the trained Q network t ;

[0038] S45, based on the selected optimal action a t Execute corresponding security policy adjustments, including power outages, alarms, and adjustment of power load distribution. The execution results in a new state s t+1 and instant rewards t+1 ;

[0039] S46, the state s t 、Action a t 、Instant Rewards t+1 and the new state s t+1 Store it in the experience replay pool D and regularly update the Q network with new data:

[0040]

[0041] By storing state transfer samples into the experience replay pool D, and using these samples to regularly update the Q network;

[0042] S47, according to the optimized control strategy, dynamically adjust and optimize the security strategy, based on the new state s t+1 and new Q values, continuously optimizing and adjusting the system's security strategy.

[0043] Optionally, the step S5 specifically includes:

[0044] S51, inputting the real-time data X(t) collected from the multi-sensor integrated device into the quantum computing device, and the quantum computing device uses the efficient parallel computing capability of quantum computing to process the data;

[0045] S52, in the quantum computing device, using quantum state superposition and quantum entanglement characteristics to convert the sensor data into a quantum state representation ψ(X(t));

[0046] S53, anomaly detection is performed through quantum computing, and the quantum variational algorithm is used to optimize the objective function:

[0047]

[0048] Among them, θ q is the quantum circuit parameter, H is the Hamiltonian, which represents the energy function of the system, λ is the regularization parameter, X i and X j They represent the components of different sensor data respectively;

[0049] S54, convert the quantum computing results back to classical data to generate quantum computing-assisted anomaly detection indicators Q d (t);

[0050] S55, quantum computing-assisted anomaly detection indicator Q d (t) Compare with historical data H(X):

[0051]

[0052] Among them, μ(H(X)) and σ(H(X)) represent the mean and standard deviation of historical data respectively;

[0053] S56, in the built-in control unit, according to the quantum computing-assisted anomaly detection indicator Q d (t) and anomaly score A(t), predicting potential anomalies;

[0054] S57, combined with the results of quantum computing, dynamically adjust and optimize the security control strategy based on the new state s t+1 And the detection results, implement corresponding safety measures, including power off, alarm and adjustment of power load distribution.

[0055] Optionally, the step S6 specifically includes:

[0056] S61, inputting the real-time data X(t) collected from the multi-sensor integrated device into the data processing device, and the data processing device preliminarily processes and compresses the data to adapt to the transmission requirements of the 5G network;

[0057] S62, the processed and compressed real-time data X p (t) Transmission via 5G networks;

[0058] S63, receiving the transmitted data on the remote monitoring platform X p (t), and perform quantum computing on the remote monitoring platform to convert the transmitted data into a quantum state representation ψ(X p (t)):

[0059] ψ(X p (t))=∑ i α i |X i >

[0060] Among them, α i is the amplitude coefficient of the quantum state, |X i > is the ground state;

[0061] S64, using the converted quantum state data ψ(X p (t)), data analysis and anomaly detection via quantum support vector machines:

[0062]

[0063] Among them, θ q is the quantum circuit parameter, y i is the label of the i-th sample, K is the quantum kernel function, b is the bias term, (·) + Indicates the positive part value;

[0064] S65, convert the calculation results of the quantum support vector machine back to the classical data format to generate the quantum computing-assisted anomaly detection indicator Q d (t);

[0065] S66. By comparing the quantum computing-assisted anomaly detection indicator Q d (t) Further analysis and prediction with historical data H(X);

[0066] S67. The user views the system status, monitoring data and alarm information through the remote monitoring platform and mobile devices, and the platform and devices are connected to the data transmission device through an interface.

[0067] Optionally, the step S7 specifically includes:

[0068] S71, the processed data X p (t) is input into the topological data analysis unit, and anomaly detection is performed by calculating persistent homology to generate topological features T f (t);

[0069] S72, the generated topological feature T f (t) is input into the Gaussian process regression unit, and the Gaussian process is used to model and predict the topological features and calculate the anomaly detection index A. d (t);

[0070] S73, calculate the abnormal detection index A d (t) is compared with the set threshold θ. dWhen (t)>θ, the system determines that it is an abnormal situation and triggers the sound and light alarm device to sound an alarm;

[0071] S74, the sound and light alarm device sends out corresponding alarm signals according to the abnormality detection results, including audio signals S audio and light signal S light , audio signal and light signal according to A d (t) is adjusted;

[0072] S75, through the quantum communication device, the abnormal detection indicator A d (t) and alarm information transmission;

[0073] S76. Transmit the abnormal detection indicators and alarm information transmitted through the quantum communication device in real time through the 5G network to achieve remote transmission of data;

[0074] S77, the remote monitoring platform receives the transmitted alarm information I a (t), after parsing, it is displayed on the user interface, including anomaly detection indicators, sound and light alarm status, and specific anomaly description;

[0075] S78. Users view system status and history records through remote monitoring platforms and mobile devices, and perform remote operations and management based on alarm information. Remote commands are sent to the built-in control unit through the 5G network to perform corresponding operations.

[0076] Optionally, the following devices are included:

[0077] Multi-sensor integrated device: including current sensor, voltage sensor, temperature sensor, humidity sensor and smoke sensor, used to monitor current, voltage, temperature, humidity and smoke environmental parameters in real time;

[0078] Data processing device: including a preprocessing unit, a topological data analysis unit and a Gaussian process regression unit, which performs preliminary processing and noise reduction on the collected data, and performs analysis to detect anomalies;

[0079] Anomaly detection device: Comparison of anomaly detection index A d (t) and the preset threshold θ to determine whether to trigger an alarm;

[0080] Alarm device: including sound and light alarm device, audio signal S audio and light signal S light According to A d (t) value is adjusted and an alarm is sounded;

[0081] Quantum computing device: used to process and analyze large-scale sensor data in real time, and to assist in anomaly detection through quantum computing, thereby improving the speed and accuracy of detection;

[0082] Quantum communication device: used to transmit abnormal detection indicators and alarm information to ensure the reliability and stability of data transmission;

[0083] 5G communication device: used to transmit data in real time and realize remote monitoring and management;

[0084] Remote monitoring and management device: includes remote monitoring platform, mobile device application and control interface, used to receive alarm information, display system status and provide remote control function.

[0085] The beneficial effects of the present invention are:

[0086] The present invention utilizes topological data analysis and Gaussian process regression units to analyze the collected data, and can detect abnormal situations in power usage with high precision; topological data analysis extracts complex topological features in the data by calculating the persistent coherence of the data, while Gaussian process regression utilizes these features for modeling and prediction to generate anomaly detection indicators; compared with traditional data analysis methods, this method can more accurately identify abnormal situations, improve the accuracy and robustness of detection, thereby significantly reducing the probability of false alarms and missed alarms, and improving the safety and reliability of the system.

[0087] The system of the present invention has intelligent early warning and automatic control functions, and can automatically take corresponding safety measures according to the detected abnormal situation. For example, when an abnormality is detected, the system can automatically trigger the sound and light alarm device to sound an alarm, including audio signals and light signals, which will be adjusted according to the value of the abnormal detection index to reflect the severity of the abnormal situation; the system can also automatically cut off or restore power supply to prevent further damage to power equipment or avoid potential safety hazards; this intelligent automatic control reduces human intervention, improves the system response speed and accuracy, and greatly improves the safety of electricity use.

[0088] In terms of data processing, the present invention introduces an improved Neural Ordinary Differential Equation algorithm, which improves the detection accuracy of abnormal situations by modeling sensor data in continuous time and analyzing dynamic changes; combined with the reinforcement learning mechanism, the system can automatically optimize and adjust security policies according to historical data and real-time environmental changes, realize intelligent and adaptive security control, and further improve the response speed and accuracy of the system; it also combines quantum computing technology, assists in anomaly detection through quantum computing, and uses the efficient parallel computing capabilities of quantum computing to process and analyze large-scale sensor data in real time. Quantum computing technology significantly improves the speed and accuracy of anomaly detection, especially in high-interference environments, ensuring the stability and security of data transmission.

[0089] In terms of data transmission, the present invention combines quantum communication technology and 5G networks to achieve efficient and secure transmission of abnormal detection indicators and alarm information. Quantum communication technology ensures the security and integrity of data transmission through quantum key distribution and encryption, and prevents data from being tampered with or stolen during transmission; while 5G networks provide high-speed, low-latency data transmission channels, so that alarm information and monitoring data can be transmitted to remote monitoring platforms and mobile devices in real time, supporting real-time remote monitoring and management; through the combination of quantum communication and 5G networks, the reliability of data transmission and the response speed of the system are greatly improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0091] Figure 1 This is a schematic diagram of the overall structure of the safety control method of the smart power safety control cabinet based on the IoT sensor proposed in the present invention;

[0092] Figure 2 is a flow chart of data processing and anomaly detection of the present invention;

[0093] Figure 3 It is a schematic diagram of the remote monitoring and management system architecture of the present invention. DETAILED DESCRIPTION

[0094] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, which only illustrate the basic structure of the present invention in a schematic manner, and therefore only show the components related to the present invention.

[0095] refer to Figure 1-3 The safety control method of the smart power safety control cabinet based on the IoT sensor includes the following steps:

[0096] S1. Install a multi-sensor integrated device in the power system, including current sensors, voltage sensors, temperature sensors, humidity sensors and smoke sensors, to monitor various parameters of the power system and environment in real time;

[0097] S2, collects current, voltage, temperature, humidity and smoke data in real time through a multi-sensor integrated device, and transmits the data to a built-in control unit for intelligent analysis;

[0098] S3, built-in control unit uses improved Neural Ordinary Differential Equation algorithm to perform continuous time modeling and dynamic change analysis on sensor data;

[0099] S4. The improved Neural Ordinary Differential Equation algorithm is combined with the deep Q-learning algorithm to automatically optimize and adjust the security strategy according to historical data and real-time environmental changes, and realize intelligent and adaptive security control, including power outage, alarm and adjustment of power load distribution;

[0100] S5. Use quantum computing to assist anomaly detection, process and analyze large-scale sensor data in real time through the parallel computing power of quantum computing, and transmit data in interference environments;

[0101] S6. Use quantum communication and 5G networks to achieve real-time transmission of monitoring data. Users can view system status, monitoring data and alarm information through remote monitoring platforms and mobile devices, and perform remote operations and management.

[0102] S7. When an abnormal situation is detected, the system will issue an audible and visual alarm, and send the alarm information to the remote monitoring platform and the mobile devices of relevant personnel through quantum communication and 5G network;

[0103] S8. The system automatically records all monitoring data and alarm events, forms historical records, and provides data analysis functions to help users trace faults and optimize power management strategies.

[0104] In this implementation, step S3 specifically includes:

[0105] S31, collecting sensor data X(t), where X(t) includes current I(t), voltage V(t), temperature T(t), humidity H(t) and smoke concentration S(t);

[0106] S32, inputting the collected sensor data X(t) into the improved neural ordinary differential equation model:

[0107]

[0108] Where f(X(t),θ) represents the basic neural ordinary differential equation part, α is the weight coefficient used to adjust the influence of the time-dependent compensation function g(t,φ), and θ and φ are model parameters;

[0109] S33, training the improved neural ordinary differential equation model through a multi-stage optimization algorithm to determine model parameters θ and φ;

[0110] S34, using the optimized model parameters θ and φ to analyze the real-time sensor data:

[0111]

[0112] in, Represents the predicted state at time t+Δt, and calculates the dynamic changes of data through integration;

[0113] S35. Combine historical data H(X) and real-time data X(t) for comprehensive analysis to predict potential abnormal situations:

[0114]

[0115] Where P(A|X,H) is the predicted probability of anomaly, σ is the activation function, and W and b are model parameters;

[0116] S36. Based on the prediction results, the optimized model parameters θ and φ are used to continuously monitor the sensor data and analyze the dynamic changes, and the safety control strategy is updated.

[0117] In this implementation, step S4 specifically includes:

[0118] S41. Define the state s of deep Q learning t 、Action a t and the reward function r t ;

[0119] S42, define the state s t and action a t And the corresponding reward r t Input into the improved Neural ODE model;

[0120] S43, combined with the deep Q learning algorithm, by using the improved neural ordinary differential equation model to calculate the state s t and instant rewards t , the Q network parameters θ are optimized by minimizing the following loss function:

[0121]

[0122] Among them, D(s t ,a t ; θ) represents the state s t Next, perform action a t Q value, θ - is the parameter of the target Q network, γ is the discount factor, and D is the experience replay pool used to store state transition samples;

[0123] S44, apply the trained control strategy in the built-in control unit, monitor the sensor data X(t) in real time and t Choose the best action a t :

[0124]

[0125] Determine the current state s by monitoring the real-time sensor data X(t) t, and select the optimal action a based on the trained Q network t ;

[0126] S45, based on the selected optimal action a t Execute corresponding security policy adjustments, including power outages, alarms, and adjustment of power load distribution. The execution results in a new state s t+1 and instant rewards t+1 ;

[0127] S46, the state s t 、Action a t 、Instant Rewards t+1 and the new state s t+1 Store it in the experience replay pool D and regularly update the Q network with new data:

[0128]

[0129] By storing state transfer samples into the experience replay pool D, and using these samples to regularly update the Q network;

[0130] S47, according to the optimized control strategy, dynamically adjust and optimize the security strategy, based on the new state s t+1 and new Q values, continuously optimizing and adjusting the system's security strategy.

[0131] In this implementation, step S5 specifically includes:

[0132] S51, inputting the real-time data X(t) collected from the multi-sensor integrated device into the quantum computing device, and the quantum computing device uses the efficient parallel computing capability of quantum computing to process the data;

[0133] S52, in the quantum computing device, using quantum state superposition and quantum entanglement characteristics to convert the sensor data into a quantum state representation ψ(X(t));

[0134] S53, anomaly detection is performed through quantum computing, and the quantum variational algorithm is used to optimize the objective function:

[0135]

[0136] Among them, θ q is the quantum circuit parameter, H is the Hamiltonian, which represents the energy function of the system, λ is the regularization parameter, X i and X j They represent the components of different sensor data respectively;

[0137] S54, convert the quantum computing results back to classical data to generate quantum computing-assisted anomaly detection indicators Q d (t);

[0138] S55, quantum computing-assisted anomaly detection indicator Q d (t) Compare with historical data H(X):

[0139]

[0140] Among them, μ(H(X)) and σ(H(X)) represent the mean and standard deviation of historical data respectively;

[0141] S56, in the built-in control unit, according to the quantum computing-assisted anomaly detection indicator Q d (t) and anomaly score A(t), predicting potential anomalies;

[0142] S57, combined with the results of quantum computing, dynamically adjust and optimize the security control strategy based on the new state s t+1 And the detection results, implement corresponding safety measures, including power off, alarm and adjustment of power load distribution.

[0143] In this implementation, step S6 specifically includes:

[0144] S61, inputting the real-time data X(t) collected from the multi-sensor integrated device into the data processing device, and the data processing device preliminarily processes and compresses the data to adapt to the transmission requirements of the 5G network;

[0145] S62, the processed and compressed real-time data X p (t) Transmission via 5G networks;

[0146] S63, receiving the transmitted data on the remote monitoring platform X p (t), and perform quantum computing on the remote monitoring platform to convert the transmitted data into a quantum state representation ψ(X p (t)):

[0147] ψ(X p (t))=∑ i α i |X i >

[0148] Among them, α i is the amplitude coefficient of the quantum state, |X i > is the ground state;

[0149] S64, using the converted quantum state data ψ(X p (t)), data analysis and anomaly detection via quantum support vector machines:

[0150]

[0151] Among them, θq is the quantum circuit parameter, y i is the label of the i-th sample, K is the quantum kernel function, b is the bias term, (·) + Indicates the positive part value;

[0152] S65, convert the calculation results of the quantum support vector machine back to the classical data format to generate the quantum computing-assisted anomaly detection indicator Q d (t);

[0153] S66. By comparing the quantum computing-assisted anomaly detection indicator Q d (t) Further analysis and prediction with historical data H(X);

[0154] S67. The user views the system status, monitoring data and alarm information through the remote monitoring platform and mobile devices, and the platform and devices are connected to the data transmission device through an interface.

[0155] In this implementation, step S7 specifically includes:

[0156] S71, the processed data X p (t) is input into the topological data analysis unit, and anomaly detection is performed by calculating persistent homology to generate topological features T f (t);

[0157] S72, the generated topological feature T f (t) is input into the Gaussian process regression unit, and the Gaussian process is used to model and predict the topological features and calculate the anomaly detection index A. d (t);

[0158] S73, calculate the abnormal detection index A d (t) is compared with the set threshold θ. d When (t)>θ, the system determines that it is an abnormal situation and triggers the sound and light alarm device to sound an alarm;

[0159] S74, the sound and light alarm device sends out corresponding alarm signals according to the abnormality detection results, including audio signals S audio and light signal S light , audio signal and light signal according to A d (t) is adjusted;

[0160] S75, through the quantum communication device, the abnormal detection indicator A d (t) and alarm information transmission;

[0161] S76. Transmit the abnormal detection indicators and alarm information transmitted through the quantum communication device in real time through the 5G network to achieve remote transmission of data;

[0162] S77, the remote monitoring platform receives the transmitted alarm information I a (t), after parsing, it is displayed on the user interface, including anomaly detection indicators, sound and light alarm status, and specific anomaly description;

[0163] S78. Users view system status and history records through remote monitoring platforms and mobile devices, and perform remote operations and management based on alarm information. Remote commands are sent to the built-in control unit through the 5G network to perform corresponding operations.

[0164] In this embodiment, the following devices are included:

[0165] Multi-sensor integrated device: including current sensor, voltage sensor, temperature sensor, humidity sensor and smoke sensor, used to monitor current, voltage, temperature, humidity and smoke environmental parameters in real time;

[0166] Data processing device: including a preprocessing unit, a topological data analysis unit and a Gaussian process regression unit, which performs preliminary processing and noise reduction on the collected data, and performs analysis to detect anomalies;

[0167] Anomaly detection device: Comparison of anomaly detection index A d (t) and the preset threshold θ to determine whether to trigger an alarm;

[0168] Alarm device: including sound and light alarm device, audio signal S audio and light signal S light According to A d (t) value is adjusted and an alarm is sounded;

[0169] Quantum computing device: used to process and analyze large-scale sensor data in real time, and to assist in anomaly detection through quantum computing, thereby improving the speed and accuracy of detection;

[0170] Quantum communication device: used to transmit abnormal detection indicators and alarm information to ensure the reliability and stability of data transmission;

[0171] 5G communication device: used to transmit data in real time and realize remote monitoring and management;

[0172] Remote monitoring and management device: includes remote monitoring platform, mobile device application and control interface, used to receive alarm information, display system status and provide remote control function.

[0173] Embodiment 1:

[0174] In order to verify the feasibility of the present invention in implementation, the present invention is applied in a large commercial office building, and the safety and reliability of the power system are of vital importance. There are many electrical equipment in the office building, including computers, servers, lighting equipment, air-conditioning systems, etc. If a power system failure occurs, it will not only affect office efficiency, but may also cause safety hazards such as fire; therefore, how to achieve real-time monitoring and intelligent management of the power system has become the primary task of the property management department; in order to improve the safety and reliability of the power system, the property management department decided to introduce the safety control method device and method of the smart power safety control cabinet based on the Internet of Things sensor of the present invention. The device realizes comprehensive real-time monitoring of the power system and environmental parameters in the office building by integrating current, voltage, temperature, humidity and smoke sensors.

[0175] A multi-sensor integrated device is installed in the main power distribution cabinet in the office building. Current sensors and voltage sensors monitor the operating status of the power system in real time, temperature sensors and humidity sensors monitor the environmental conditions between the power distribution cabinet and the equipment, and smoke sensors can detect possible fire hazards in time. The data of all sensors are transmitted to the data processing device through the wireless communication module.

[0176] The data processing device performs preprocessing and noise reduction on the collected data to generate preprocessed data. Then, these data are input into the topological data analysis unit to generate topological features by calculating persistent homology. Then, the topological features are input into the Gaussian process regression model for modeling and prediction to generate anomaly detection indicators; in order to further improve the accuracy of anomaly detection, the present invention introduces an improved Neural Ordinary Differential Equation algorithm, combined with a deep learning algorithm to perform continuous time modeling and dynamic change analysis on sensor data. The system can automatically optimize and adjust security policies based on historical data and real-time environmental changes to achieve intelligent and adaptive security control.

[0177] Once an abnormality is detected, the control unit will automatically trigger the sound and light alarm device to sound an alarm, including audio and light signals. These signals are adjusted according to the value of the abnormality detection index to reflect the severity of the abnormality. At the same time, the system will transmit the alarm information and abnormality detection indicators to the remote monitoring platform and the mobile devices of the property management personnel in real time through the quantum communication module and 5G network.

[0178] Property managers can view system status, monitoring data and alarm information in real time through remote monitoring platforms and mobile devices, and perform remote operations and management. For example, when the system detects that the temperature of a power distribution cabinet has risen abnormally, property managers can immediately view the specific situation through mobile devices and remotely instruct power off to prevent overheating of equipment and cause fire. At the same time, the system automatically records all monitoring data and alarm events to form historical records for subsequent analysis and fault tracing.

[0179] In actual application, the smart power safety control cabinet device significantly improves the safety and reliability of the office building's power system. The following is a data comparison of an office building before and after the application of the device of the present invention: From January 2023 to June 2023, the office building had a total of 5 power system failures, including 2 power outages caused by equipment overheating and 3 tripping events caused by current overload. After the introduction of the device of the present invention, from July 2023 to December 2023, the office building only had 1 minor current overload event, and did not cause any equipment damage or safety hazards.

[0180] The specific data are as follows: the number of power outages caused by equipment overheating was reduced from 2 times to 0 times, the number of tripping events caused by current overload was reduced from 3 times to 1 time, the average response time for each fault was shortened from 30 minutes to 5 minutes, and the equipment repair cost after the fault was reduced from an average of 5,000 yuan per time to 1,000 yuan. The total power system maintenance cost was reduced by about 40%.

[0181] Table 1 Statistics of office building power system failures

[0182]

[0183]

[0184] Table 2 Data comparison before and after the introduction of smart power safety control cabinet device in office building

[0185]

[0186] It can be seen from the above data that the statistics of power system failures in the office building in the first and second half of 2023 are recorded in detail, showing the improvement of the system in terms of the number of failures, response time, repair costs, maintenance costs and equipment damage. From January to June 2023, there were 5 power system failures, including 2 power outages caused by equipment overheating and 3 tripping events caused by current overload. From July to December 2023, only 1 minor current overload event occurred, and no equipment damage or safety hazards were caused. Before the introduction of the device of the present invention, the average response time was 30 minutes; after the introduction, the response time was shortened to 5 minutes. The cost of equipment repair after the failure was reduced from an average of 5,000 yuan per time to 1,000 yuan. The total power system maintenance cost was significantly reduced, and the equipment damage was also improved from severe / moderate to slight / none.

[0187] The detailed comparative data before and after the introduction of the device of the present invention are further displayed, covering aspects such as equipment overheating, fault response time, repair costs, maintenance costs, and abnormality detection accuracy. The number of power outages caused by equipment overheating was reduced from 2 to 0, the number of tripping events caused by current overload was reduced from 3 to 1, the average response time for each fault was shortened from 30 minutes to 5 minutes, and the equipment repair cost after each fault was reduced from an average of 5,000 yuan per time to 1,000 yuan. The total power system maintenance cost was reduced from high to low, and the total maintenance cost decreased by 40%. The accuracy of abnormality detection was increased from 85% to 95%, the alarm response time was shortened from 30 minutes to 5 minutes, the data transmission reliability was increased from 80% to 99%, the user satisfaction was increased from 60% to 95%, the equipment operation stability was improved from medium to high, and the frequency of faults was reduced from high to low.

[0188] Through the summary of the table data, it can be clearly seen that after the introduction of the intelligent power safety control cabinet device of the present invention, the office building power system has been significantly improved in terms of safety, reliability and management efficiency. The reduction in the number of failures, the shortened response time, the reduced repair costs and the reduced maintenance costs all indicate that the device of the present invention has effectively improved the abnormal detection and response capabilities of the power system, reduced the number of failures and maintenance costs, and improved the safety and management efficiency of the office building. These improvements not only improve the operating stability and user satisfaction of the equipment, but also significantly reduce the frequency of failures, demonstrating the practicality and effectiveness of the present invention.

[0189] In summary, the smart power safety control cabinet of the present invention has shown significant beneficial effects in practical applications and solved many problems in the safety management of traditional power systems. By introducing advanced sensing technology, data analysis algorithms, quantum computing and communication technology, the present invention significantly improves the safety, reliability and management efficiency of the power system and has broad application prospects.

[0190] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A safety control method for a smart power safety control cabinet based on IoT sensors, characterized in that: The steps include: S1. Install a multi-sensor integrated device in the power system, including current sensors, voltage sensors, temperature sensors, humidity sensors and smoke sensors, to monitor various parameters of the power system and environment in real time; S2, collects current, voltage, temperature, humidity and smoke data in real time through a multi-sensor integrated device, and transmits the data to a built-in control unit for intelligent analysis; S3, built-in control unit uses improved Neural Ordinary Differential Equation algorithm to perform continuous time modeling and dynamic change analysis on sensor data; S4. The improved Neural Ordinary Differential Equation algorithm is combined with the deep Q-learning algorithm to automatically optimize and adjust the security strategy according to historical data and real-time environmental changes, and realize intelligent and adaptive security control, including power outage, alarm and adjustment of power load distribution; S5. Use quantum computing to assist anomaly detection, process and analyze large-scale sensor data in real time through the parallel computing power of quantum computing, and transmit data in interference environments; The step S5 specifically includes: S51, inputting the sensor data X(t) collected from the multi-sensor integrated device into the quantum computing device, and the quantum computing device uses the efficient parallel computing capability of quantum computing to process the data; S52, in the quantum computing device, using quantum state superposition and quantum entanglement characteristics to convert the sensor data into a quantum state representation ψ(X(t)); S53, anomaly detection is performed through quantum computing, and the quantum variational algorithm is used to optimize the objective function: Among them, θ q is the quantum circuit parameter, H is the Hamiltonian, which represents the energy function of the system, λ is the regularization parameter, X i and X j They represent the components of different sensor data respectively; S54, convert the quantum computing results back to classical data to generate quantum computing-assisted anomaly detection indicators Q d (t); S6. Use quantum communication and 5G networks to achieve real-time transmission of monitoring data. Users can view the power system status, monitoring data and alarm information through remote monitoring platforms and mobile devices, and perform remote operations and management. S7. The power consumption system will issue an audible and visual alarm when an abnormal situation is detected, and send the alarm information to the remote monitoring platform and the mobile devices of relevant personnel through quantum communication and 5G network; S8. The power consumption system automatically records all monitoring data and alarm events, forms historical records, and provides data analysis functions to help users trace faults and optimize power consumption management strategies.

2. The safety control method of the smart power safety control cabinet based on the IoT sensor according to claim 1 is characterized in that: The step S3 specifically includes: S31, collecting sensor data X(t), where X(t) includes current I(t), voltage V(t), temperature T(t), humidity H(t) and smoke concentration S(t); S32, inputting the collected sensor data X(t) into the improved neural ordinary differential equation model: Where f(X(t),θ) represents the basic neural ordinary differential equation part, α is the weight coefficient used to adjust the influence of the time-dependent compensation function g(t,φ), and θ and φ are model parameters; S33, training the improved neural ordinary differential equation model through a multi-stage optimization algorithm to determine model parameters θ and φ; S34, using the optimized model parameters θ and φ to analyze the real-time sensor data: in, Represents the predicted state at time t+Δt, and calculates the dynamic changes of data through integration; S35. Combine historical data H(X) and sensor data X(t) for comprehensive analysis to predict potential abnormal situations: Where P(A|X,H) is the predicted probability of anomaly, σ is the activation function, and W and b are model parameters; S36. Based on the prediction results, the optimized model parameters θ and φ are used to continuously monitor the sensor data and analyze the dynamic changes, and the safety control strategy is updated.

3. The safety control method of the smart power safety control cabinet based on the IoT sensor according to claim 2 is characterized in that: The step S4 specifically includes: S41. Define the state s of deep Q learning t 、Action a t and instant rewards t ; S42, define the state s t and action a t And the corresponding instant reward r t Input into the improved Neural ODE model; S43, combined with the deep Q learning algorithm, by using the improved neural ordinary differential equation model to calculate the state s t and instant rewards t , the Q network parameters θ are optimized by minimizing the following loss function: Among them, Q(s t ,a t ; θ) represents the state s t Next, perform action a t Q value, θ - is the parameter of the target Q network, γ is the discount factor, and D is the experience replay pool used to store state transition samples; S44, apply the trained control strategy in the built-in control unit, monitor the sensor data X(t) in real time and t Choose the best action a t : π(s t )=arg max a Q(s t ,a;i); Determine the current state s by monitoring the real-time sensor data X(t) t , and select the optimal action a based on the trained Q network t ; S45, based on the selected optimal action a t Execute corresponding security policy adjustments, including power outages, alarms, and adjustment of power load distribution. The execution results in a new state s t+1 and instant rewards t+1 ; S46, the state s t 、Action a t , Instant Rewards t+1 and the new state s t+1 Store it in the experience replay pool D and regularly update the Q network with new data: By storing state transfer samples into the experience replay pool D, and using these samples to regularly update the Q network; S47, according to the optimized control strategy, dynamically adjust and optimize the security strategy, based on the new state s t+1 And new Q values, continuously optimize and adjust the safety strategy of the power system.

4. The safety control method of the smart power safety control cabinet based on the IoT sensor according to claim 1 is characterized in that: The step S5 specifically includes: S55, quantum computing-assisted anomaly detection indicator Q d (t) Compare with historical data H(X): Among them, μ(H(X)) and σ(H(X)) represent the mean and standard deviation of historical data respectively; S56, in the built-in control unit, according to the quantum computing-assisted anomaly detection indicator Q d (t) and anomaly score A(t), predicting potential anomalies; S57, combined with the results of quantum computing, dynamically adjust and optimize the security control strategy based on the new state s t+1 And the detection results, implement corresponding safety measures, including power off, alarm and adjustment of power load distribution.

5. The safety control method of the smart power safety control cabinet based on the IoT sensor according to claim 1 is characterized in that: The step S6 specifically includes: S61, inputting the sensor data X(t) collected from the multi-sensor integration device into the data processing device, and the data processing device preliminarily processes and compresses the data to adapt to the transmission requirements of the 5G network; S62, the processed and compressed sensor data X p (t) Transmission via 5G networks; S63, receiving the transmitted data on the remote monitoring platform X p (t), and perform quantum computing on the remote monitoring platform to convert the transmitted data into a quantum state representation ψ(X p (t)): ψ(X p (t))=∑ i a i |X i >; Among them, α i is the amplitude coefficient of the quantum state, |X i > is the ground state; S64, using the converted quantum state data ψ(X p (t)), data analysis and anomaly detection via quantum support vector machines: Among them, θ q is the quantum circuit parameter, y i is the label of the i-th sample, K is the quantum kernel function, b is the bias term, (·) + Indicates the positive part value; S65, convert the calculation results of the quantum support vector machine back to the classical data format to generate the quantum computing-assisted anomaly detection indicator Q' d (t); S66. By comparing the quantum computing-assisted anomaly detection indicator Q' d (t) Further analysis and prediction with historical data H(X); S67. Users view the power system status, monitoring data and alarm information through the remote monitoring platform and mobile devices. The platform and devices are connected to the data transmission device through an interface.

6. The safety control method of the smart power safety control cabinet based on the IoT sensor according to claim 5 is characterized in that: The step S7 specifically includes: S71, the processed data X p (t) is input into the topological data analysis unit, and anomaly detection is performed by calculating persistent homology to generate topological features T f (t); S72, the generated topological feature T f (t) is input into the Gaussian process regression unit, and the Gaussian process is used to model and predict the topological features and calculate the anomaly detection index A. d (t); S73, calculate the abnormal detection index A d (t) is compared with the set threshold θ. d When (t)>θ, it is determined to be an abnormal situation and the sound and light alarm device is triggered to sound an alarm; S74, the sound and light alarm device sends out corresponding alarm signals according to the abnormality detection results, including audio signals S audio and light signal S light , audio signal and light signal according to A d (t) is adjusted; S75, through the quantum communication device, the abnormal detection indicator A d (t) and alarm information transmission; S76. Transmit the abnormal detection indicators and alarm information transmitted through the quantum communication device in real time through the 5G network to achieve remote transmission of data; S77, the remote monitoring platform receives the transmitted alarm information I a (t), after parsing, it is displayed on the user interface, including anomaly detection indicators, sound and light alarm status, and specific anomaly description; S78. Users can view the status and history of the power system through the remote monitoring platform and mobile devices, and perform remote operations and management based on alarm information. Remote commands are sent to the built-in control unit through the 5G network to perform corresponding operations.

Images