Transition method for intelligent transformation of transformer substation protection equipment

Abstract

A transition method for intelligent transformation of transformer substation protection equipment comprises the step of additionally arranging an intelligent interface device, and comprises the following steps: Step 1, a bus is powered off, and a previous bus protection device is replaced with an intelligent bus protection device capable of receiving and sending optical digital signals; cables of all the unmodified interval circuit breaker protection devices and the operation box connected to the bus are connected to a switching value electric signal end of the intelligent interface device; 3, the transformed intelligent bus protection device is put into operation, and the bus is powered on for operation; step 4, power is cut off for the intervals one by one, and intelligent transformation is carried out; and Step 5, after all the unmodified interval circuit breaker protection devices and the operation box are replaced, the intelligent interface device is dismounted. By additionally arranging transition equipment, the cost is low, and implementation is convenient; after intelligent bus protection transformation is completed, bus protection functions can be synchronously and completely put into use, the power failure time of the bus is short, and the reliability and safety of operation of the transformer substation are guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of transformer substation reconstruction, and in particular to a transition method for intelligent reconstruction of transformer substation protection equipment. Background Art

[0002] Relay protection equipment is an indispensable electrical secondary equipment in both conventional substations and smart substations. The switch quantity signal transmission methods of protection devices in conventional substations and smart substations are different. Conventional relay protection uses cables to transmit switch quantity signals, while smart relay protection uses optical fibers to transmit switch quantity signals. There is no doubt that optical fiber digitization is the direction of equipment transformation.

[0003] Taking the substation with two-thirds wiring as an example, the busbar protection needs to receive the "failure through busbar tripping" input information of all circuit breakers connected to the busbar, and send the trip command output information to each bay; the conventional busbar protection of conventional substations receives and sends switch quantity information by cable, such as Figure 1 Intelligent busbar protection can only receive and send optical digital signals and is not compatible with the reception and transmission of conventional cable switch quantities and optical digital signals at the same time.

[0004] At present, conventional substations in operation in the power industry are gradually undergoing intelligent transformation. Usually, the whole substation is shut down. The circuit breaker protection, operation box, and busbar protection of each compartment are synchronously transformed and replaced with intelligent circuit breaker protection, intelligent terminal, and intelligent busbar protection. These intelligent devices all use optical digital signal output switch quantity, and there is no obstacle. However, this transformation method requires the whole station to be shut down, and the substation circuit breakers, transformers, and lines cannot operate. During the transformation, the reliability of power transmission and load power consumption is greatly reduced, and additional power loss may also be caused. If the whole station is not shut down, but the intervals are shut down in turn, when a certain interval is transformed into intelligent protection, the untransformed conventional busbar protection cannot transmit information with the intelligent protection of the transformed interval; on the contrary, if the busbar is firstly transformed into intelligent busbar protection, in order to ensure information transmission, all the circuit breaker intervals connected to the busbar must be synchronously transformed into intelligent protection, and the conventional protection of each incoming and outgoing line connected to these circuit breaker intervals cannot transmit information normally, and it is not feasible to extend the power outage to the whole station.

[0005] In view of the shortcomings of the existing transformation methods, it is necessary to design a method for intelligent transformation of conventional substation protection equipment to achieve rotating power outages in each interval and carry out intelligent transformation of protection equipment. Summary of the invention

[0006] The technical problem to be solved by the present invention is to provide a transition method for intelligent transformation of transformer substation protection equipment, so as to realize the intelligent transformation of protection equipment while realizing the rotational power outage of each interval.

[0007] To solve the above technical problems, the technical solution adopted by the present invention is as follows:

[0008] A transition method for the intelligent transformation of substation protection equipment, including adding an intelligent interface device for converting optical digital switch signals and cable switch signals, and the intelligent interface device is arranged on the bus protection panel. The transformation steps include:

[0009] Step1. Power off the bus, replace the previous bus protection device with an intelligent bus protection device capable of receiving and sending optical digital signals, and configure an intelligent interface device on the bus configuration panel. Connect the optical signal cable of the intelligent bus protection device to the optical digital signal terminal of the intelligent interface device.

[0010] Step2. After the intelligent interface device is installed, connect the cables of all untransformed interval breaker protection devices and operation boxes connected to the bus to the switch quantity electrical signal terminals of the intelligent interface device.

[0011] Step3. Put the transformed intelligent bus protection device into operation, and the bus is powered on and operated.

[0012] Step4. Power off each interval one by one for intelligent transformation. After the breaker of interval A is powered off, interval A is transformed. Connect the transformed breaker protection and intelligent terminal of interval A, that is, the transformed intelligent interval breaker protection device and operation box, to the intelligent bus protection device through optical fiber, and then interval A is powered on again. Other intervals are transformed in the same process as interval A until all untransformed interval breaker protection devices and operation boxes are replaced.

[0013] Step5. After all untransformed interval breaker protection devices and operation boxes are replaced, remove the intelligent interface device.

[0014] In the above Step2, the intelligent interface device converts the switch quantity signals of all untransformed interval breaker protection devices and operation boxes connected to the bus into digital quantity signals and sends them to the intelligent bus protection device. The optical digital signals sent by the intelligent bus protection device are converted into switch quantity signals through the intelligent interface device and then sent to the untransformed interval breaker protection devices and operation boxes.

[0015] The above intelligent interface device includes a PLC controller and an optoelectronic conversion module. The PLC controller is used to receive the switch quantity signals from the untransformed interval breaker protection devices and operation boxes, and convert them into optical digital signals through the optoelectronic conversion module and send them to the intelligent bus protection device. At the same time, the PLC controller is also used to receive the optical digital signals sent by the intelligent bus protection device, and convert them into switch quantity signals through the optoelectronic conversion module and send them to the untransformed interval breaker protection devices and operation boxes.

[0016] The above PLC controller is also used to analyze the status of the received digital signals. When an abnormal status is detected, the PLC controller can trigger emergency response measures, including sending an alarm signal or automatically cutting off the relevant circuit.

[0017] The above intelligent interface device is built-in with a processor and a storage module, and integrates a sensor module. The processor and the storage module are used for the artificial intelligence algorithm with built-in self-learning ability and storing historical data. The sensor module is used to monitor in real time parameters such as signal transmission delay, noise, and voltage fluctuation. The intelligent interface device can automatically adjust the signal conversion parameters according to the operation status of the substation, optimize the signal transmission efficiency, and form a signal transmission optimization model based on the artificial intelligence algorithm with self-learning ability. The specific process is as follows:

[0018] Using the long short-term memory network LSTM as the core algorithm to model and predict parameters such as delay and noise in signal transmission;

[0019] Establishment of the signal transmission optimization model:

[0020] 1) Data collection and preprocessing:

[0021] Collect historical signal transmission data, including:

[0022] Signal delay D(t), noise level N(t), voltage fluctuation V(t), and signal packet loss rate L(t);

[0023] Perform normalization processing on the data:

[0024]

[0025] where X is the original data, and Xmin and Xmax are the minimum and maximum values of the data respectively;

[0026] 2) Model training

[0027] Input data: time series data {D(t), N(t), V(t), L(t)};

[0028] Output data: optimized signal transmission parameters {P(t)}, such as sampling rate and filtering intensity;

[0029] 3) Model optimization: Use the Adam optimizer to train the model, set the learning rate, and during the training process, adopt the early stopping method Early Stopping to prevent overfitting;

[0030] 4) Model Deployment: Deploy the trained LSTM model to the intelligent interface device to collect signal transmission data in real time and generate optimization strategies.

[0031] The above intelligent interface device uses a reinforcement learning algorithm to train the intelligent interface device to automatically select the optimal signal processing strategy in different operating environments. The specific process is as follows:

[0032] Adopt the Deep Q-Network (DQN) as the core algorithm to train the intelligent interface device to automatically select the optimal signal processing strategy in different operating environments;

[0033] Establishment of the reinforcement learning algorithm:

[0034] 1) Define the state space, action space, and reward function:

[0035] State space S: Includes signal delay D(t), noise level N(t), voltage fluctuation V(t), and signal packet loss rate L(t);

[0036] Action space A: Adjust the sampling rate, adjust the filtering intensity;

[0037] Reward function R:

[0038] R(t) = -αD(t) - βN(t) - γL(t);

[0039] Among them, α, β, and γ are weight coefficients used to balance the impacts of delay, noise, and packet loss rate;

[0040] 2) Construct the Deep Q-Network:

[0041] Input layer: The input dimension is 4, corresponding to 4 state features;

[0042] Hidden layer: Contains two fully connected layers, with 128 neurons in each layer;

[0043] Output layer: The output dimension is 4, corresponding to 4 actions;

[0044] Loss function: Adopt the Mean Squared Error (MSE) as the loss function:

[0045]

[0046] Among them, Q pred (s,a) is the predicted Q value, and Q target (s,a) is the target Q value;

[0047] 3) Training:

[0048] Experience replay: Store historical states, actions, rewards, and the next state for training;

[0049] Target network: The target network is used to calculate the target Q value to reduce fluctuations during the training process;

[0050] Q value update formula:

[0051] Q target (s,a) = R(t) + γmax a′ Q(s′,a′);

[0052] where γ is the discount factor, s′ is the next state, and a′ is the next action;

[0053] 4) Policy selection:

[0054] The ε-greedy policy is adopted to select actions:

[0055]

[0056] where ∈ is the exploration rate, with an initial value of 1 and gradually decaying;

[0057] 5) Model deployment: The trained DQN model is deployed into the intelligent interface device to collect status data in real time and select the optimal actions.

[0058] The above-mentioned intelligent interface device adopts multi-channel parallel signal processing technology, supports simultaneous processing of digital input signals of multiple intervals, and improves the signal processing speed and system response ability.

[0059] The above-mentioned multi-channel parallel signal processing technology is implemented through a multi-core processor or FPGA.

[0060] The above-mentioned intelligent interface device has a fault self-healing function, can automatically recover or switch to a standby module when a fault is detected, and the fault self-healing function is realized through redundant design and fault detection algorithms to ensure that the device can automatically switch to a standby module during a fault.

[0061] The above-mentioned intelligent interface device also automatically repairs software or configuration errors through a self-healing algorithm to reduce manual intervention.

[0062] A transition method for the intelligent transformation of substation protection equipment provided by the present invention has the following beneficial effects: 1. By adding a transition device as an interface conversion device between conventional protection and intelligent protection, the cost is low and the implementation is convenient;

[0063] 2. After the intelligent bus protection transformation is completed, the bus protection function can be fully synchronized and put into operation, and the bus power outage time is short, ensuring the reliability and safety of the substation operation;

[0064] 3. This method enables the intelligent transformation of conventional power stations without a full power outage of the whole station, improves the reliability of power generation and supply, and avoids possible economic losses of electricity.

[0065] 4. The intelligent interface device can dynamically adjust the signal conversion parameters according to the actual operation status of the substation, significantly improving the efficiency and stability of signal transmission. The intelligent interface device has the ability of self-learning, can analyze problems such as delay and noise in signal transmission in real time, and automatically optimize the signal processing strategy, reducing the need for manual intervention. In a complex operating environment, the intelligent interface device can adaptively adjust to ensure the high efficiency and reliability of signal transmission, improving the overall operating efficiency of the substation.

[0066] 5. Through multi-channel parallel signal processing technology, the intelligent interface device can simultaneously process the digital input signals of multiple intervals, significantly improving the throughput of signal processing, ensuring the real-time transmission of key signals (such as trip commands), avoiding signal congestion or loss, and improving the response speed of the system. During the large-scale transformation of the substation, the multi-channel parallel processing technology can significantly reduce the delay of signal processing, ensuring the smooth progress of the transformation process.

[0067] 6. The intelligent interface device has a fault self-healing function, can automatically recover or switch to a standby module when a fault is detected, significantly improving the reliability and operating stability of the device. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] The present invention will be further described below in conjunction with the drawings and embodiments:

[0069] Figure 1 It is a wiring schematic diagram of a conventional bus protection device before transformation;

[0070] Figure 2 It is a wiring schematic diagram of an intelligent bus protection device with an intelligent interface device added;

[0071] Figure 3 It is a wiring schematic diagram of the intelligent bus protection device during the transformation process;

[0072] Figure 4 It is a wiring schematic diagram of the intelligent bus protection device after the transformation is completed.

[0073] Among them: intelligent interface device 1, intelligent bus protection device 2, bus protection device 3, digital input signals of non-transformed interval breaker protection device and operation box 4, digital input signals of transformed intelligent interval breaker protection device and operation box 5. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0074] The technical solution of the present invention will be described in detail below in conjunction with the drawings and embodiments.

[0075] A transition method for the intelligent transformation of substation protection equipment includes adding an intelligent interface device 1. The intelligent interface device 1 is used for the conversion between optical digital input signals and cable digital input signals. The intelligent interface device 1 is arranged on the bus protection panel. The transformation steps include:

[0076] Step 1. De-energize the busbar, replace the previous busbar protection device 3 with an intelligent busbar protection device 2 capable of receiving and transmitting optical digital signals, and configure an intelligent interface device 1 on the busbar configuration panel. Connect the optical signal cable of the intelligent busbar protection device 2 to the optical digital signal terminal of the intelligent interface device 1;

[0077] Step 2. After installing the intelligent interface device 1, connect the cables of all unmodified interval circuit breaker protection devices and operating boxes 4 connected to the busbar to the switch quantity signal terminals of the intelligent interface device 1;

[0078] Step 3. Put the modified intelligent busbar protection device 2 into operation, and energize and operate this busbar;

[0079] Step 4. De-energize each interval one by one for intelligent transformation. After the circuit breaker of interval A is de-energized, interval A is transformed. After the circuit breaker protection and intelligent terminal after the transformation of interval A, that is, the modified intelligent interval circuit breaker protection device and operating box 5 are connected to the intelligent busbar protection device 2 through optical fibers, and then interval A is powered on again. Other intervals are transformed in the same process as interval A in turn until all unmodified interval circuit breaker protection devices and operating boxes 4 are replaced;

[0080] Step 5. After all unmodified interval circuit breaker protection devices and operating boxes 4 are replaced, remove the intelligent interface device 1.

[0081] In the above Step 2, the intelligent interface device 1 converts the switch quantity signals of all unmodified interval circuit breaker protection devices and operating boxes 4 connected to the busbar into digital quantity signals and sends them to the intelligent busbar protection device 2. The optical digital signals sent by the intelligent busbar protection device 2 are converted into switch quantity signals through the intelligent interface device 1 and then sent to the unmodified interval circuit breaker protection devices and operating boxes 4.

[0082] The above intelligent interface device 1 includes a PLC controller and an optoelectronic conversion module. The PLC controller is used to receive the switch quantity signals from the unmodified interval circuit breaker protection devices and operating boxes 4, and convert them into optical digital signals through the optoelectronic conversion module and send them to the intelligent busbar protection device 2; at the same time, the PLC controller is also used to receive the optical digital signals sent by the intelligent busbar protection device 2, and convert them into switch quantity signals through the optoelectronic conversion module and send them to the unmodified interval circuit breaker protection devices and operating boxes 4.

[0083] The above PLC controller is also used to perform status analysis on the received switch quantity signals. When an abnormal state is detected, the PLC controller can trigger emergency response measures, including sending an alarm signal or automatically cutting off relevant circuits.

[0084] The above intelligent interface device 1 has a built-in processor and storage module, and integrates a sensor module. The processor and storage module are used to build an artificial intelligence algorithm with self-learning ability and store historical data. The sensor module is used to monitor parameters such as signal transmission delay, noise, and voltage fluctuation in real time. The intelligent interface device 1 can automatically adjust signal conversion parameters according to the operating state of the substation, optimize signal transmission efficiency, and form a signal transmission optimization model based on the artificial intelligence algorithm with self-learning ability. The specific process is as follows:

[0085] The long short-term memory network (LSTM) is used as the core algorithm to model and predict parameters such as delay and noise in signal transmission;

[0086] Establishment of the signal transmission optimization model:

[0087] 1) Data collection and preprocessing:

[0088] Collect historical signal transmission data, including:

[0089] Signal delay D(t), noise level N(t), voltage fluctuation V(t), and signal packet loss rate L(t);

[0090] Normalize the data:

[0091]

[0092] where X is the original data, and Xmin and Xmax are the minimum and maximum values of the data respectively;

[0093] 2) Model training

[0094] Input data: time series data {D(t), N(t), V(t), L(t)};

[0095] Output data: optimized signal transmission parameters {P(t)}, such as sampling rate and filtering intensity;

[0096] LSTM model structure:

[0097] Input layer: the input dimension is 4 (corresponding to 4 features);

[0098] LSTM layer: contains 128 hidden units;

[0099] Fully connected layer: output the optimized signal transmission parameters;

[0100] Loss function: the mean squared error (MSE) is used as the loss function:

[0101]

[0102] Among them, P pred (t) is the model prediction value, and P true (t) is the actual value;

[0103] 3) Model optimization: Use the Adam optimizer to train the model, set the learning rate, and during the training process, adopt the early stopping method Early Stopping to prevent overfitting;

[0104] 4) Model deployment: Deploy the trained LSTM model to the intelligent interface device to collect signal transmission data in real time and generate optimization strategies.

[0105] The above intelligent interface device 1 uses the reinforcement learning algorithm to train the intelligent interface device 1 to automatically select the optimal signal processing strategy in different operating environments. The implementation steps are as follows:

[0106] During the operation of the device, collect signal transmission data in real time (such as delay, noise, packet loss rate, etc.).

[0107] Input the data into the AI model to generate signal processing optimization strategies (such as adjusting the sampling rate, increasing the filtering intensity, etc.).

[0108] According to the optimization strategy, dynamically adjust the working parameters of the signal processing module.

[0109] Example: When it is detected that the signal delay in a certain interval is relatively high, the AI model will automatically adjust the signal processing priority of this channel to ensure the real-time transmission of key signals.

[0110] The specific process is as follows:

[0111] Adopt the deep Q-network DQN as the core algorithm to train the intelligent interface device to automatically select the optimal signal processing strategy in different operating environments;

[0112] Establishment of the reinforcement learning algorithm:

[0113] 1) Define the state space, action space, and reward function:

[0114] State space S: includes signal delay D(t)D(t), noise level N(t)N(t), voltage fluctuation V(t)V(t), and signal packet loss rate L(t)L(t);

[0115] Action space A: adjust the sampling rate (such as increasing by 10%, decreasing by 10%), adjust the filtering intensity (such as increasing by 20%, decreasing by 20%);

[0116] Reward function R:

[0117] R(t) = -αD(t) - βN(t) - γL(t);

[0118] Among them, α, β, and γ are weight coefficients used to balance the impacts of delay, noise, and packet loss rate;

[0119] 2) Construct a deep Q-network:

[0120] Input layer: The input dimension is 4, corresponding to 4 state features;

[0121] Hidden layer: It contains two fully connected layers, with 128 neurons in each layer;

[0122] Output layer: The output dimension is 4, corresponding to 4 actions;

[0123] Loss function: The mean squared error MSE is used as the loss function:

[0124]

[0125] Among them, Q pred (s,a) is the predicted Q value, and Q target (s,a) is the target Q value;

[0126] 3) Training:

[0127] Experience replay: Store historical states, actions, rewards, and next states for training;

[0128] Target network: Use the target network to calculate the target Q value to reduce fluctuations during the training process;

[0129] Q value update formula:

[0130] Q target (s,a) = R(t) + γmax a′ Q(s′,a′);

[0131] Among them, γ is the discount factor, s′ is the next state, and a′ is the next action;

[0132] 4) Policy selection:

[0133] Adopt the ε-greedy policy to select actions:

[0134]

[0135] Among them, ∈ is the exploration rate, with an initial value of 1 and gradually decaying;

[0136] 5) Model deployment: Deploy the trained DQN model to the intelligent interface device to collect state data in real time and select the optimal actions.

[0137] The above intelligent interface device 1 adopts multi-channel parallel signal processing technology, supports the simultaneous processing of digital input signals at multiple intervals, and improves the signal processing speed and system response ability.

[0138] The above multi-channel parallel signal processing technology is implemented through a multi-core processor or FPGA, ensuring the real-time transmission of key signals and avoiding signal congestion or loss.

[0139] The above intelligent interface device 1 has a self-healing function for faults, can automatically recover or switch to a standby module when a fault is detected, and the self-healing function for faults is achieved through redundant design and fault detection algorithms, ensuring that the device can automatically switch to a standby module when a fault occurs.

[0140] The above intelligent interface device 1 also automatically repairs software or configuration errors through a self-healing algorithm, reducing manual intervention.

[0141] Embodiment 1:

[0142] During the intelligent transformation of a 220 kV substation, the transition method described in the present invention is adopted. First, after the bus is de-energized, the original bus protection device is replaced with an intelligent bus protection device, and an intelligent interface device is installed. The intelligent interface device uses an optoelectronic conversion module with the model SIC-2000, which is produced by Beijing Youchuang Xintong Technology Co., Ltd. This module can convert the cable digital input signal into an optical digital signal and vice versa.

[0143] During the transformation process, the intelligent interface device successfully converts the digital input signals of the circuit breaker protection device and the operating box of the untransformed interval into optical digital signals and transmits them to the intelligent bus protection device. At the same time, the optical digital signals sent by the intelligent bus protection device are also converted into digital input signals through the intelligent interface device and transmitted to the circuit breaker protection device and the operating box of the untransformed interval.

[0144] During the transformation process, the PLC controller performs real-time status analysis on the received digital input signals. When an abnormal status is detected, the PLC controller automatically triggers emergency response measures, sends an alarm signal, and cuts off the relevant circuit, ensuring the safety of the transformation process.

[0145] Embodiment 2:

[0146] During the intelligent transformation of another 110 kV substation, the transition method described in the present invention is also adopted. The intelligent interface device uses an optoelectronic conversion module with the model SIC-1000. This module performed excellently during the transformation process and successfully achieved the bidirectional conversion of cable digital input signals and optical digital signals.

[0147] During the transformation process, the PLC controller conducts real-time status analysis on the received digital signals. When an abnormal status is detected, the PLC controller automatically triggers emergency response measures, issues an alarm signal, and cuts off the relevant circuits, ensuring the safety of the transformation process.

Claims

1. A transition method for intelligent transformation of substation protection equipment, characterized in that: The invention comprises adding an intelligent interface device (1), the intelligent interface device (1) is used for converting optical digital switch quantity and cable switch quantity signals, and the intelligent interface device (1) is arranged on a busbar protection plate. The transformation steps include: Step 1: When the busbar is powered off, the previous busbar protection device (3) is replaced with an intelligent busbar protection device (2) capable of receiving and sending optical digital signals, and an intelligent interface device (1) is configured on the busbar configuration panel, and the optical signal cable of the intelligent busbar protection device (2) is connected to the optical digital signal terminal of the intelligent interface device (1); Step 2, after the intelligent interface device (1) is installed, the cables of all the unmodified bay circuit breaker protection devices and the operation box (4) connected to the busbar are connected to the switch quantity electrical signal terminal of the intelligent interface device (1); Step 3, putting the modified intelligent busbar protection device (2) into operation, and the busbar is powered on; Step 4, power off each bay one by one and carry out intelligent transformation. After the power off of bay A circuit breaker, bay A is transformed, and the transformed circuit breaker protection and intelligent terminal of bay A, that is, the transformed intelligent bay circuit breaker protection device and operation box (5) are connected to the intelligent busbar protection device (2) through optical fiber, and then bay A is powered on again, and other bays are transformed in turn with the same process as bay A, until all unmodified bay circuit breaker protection devices and operation boxes (4) are replaced; Step 5. After all the unmodified compartment circuit breaker protection devices and operation boxes (4) have been replaced, the intelligent interface device (1) is removed.

2. A transition method for intelligent transformation of substation protection equipment according to claim 1, characterized in that: In the above-mentioned Step 2, the intelligent interface device (1) converts the switch quantity signals of all the unmodified interval circuit breaker protection devices and the operation box (4) connected to the bus into digital quantity signals and sends them to the intelligent bus protection device (2); the optical digital signals emitted by the intelligent bus protection device (2) are converted into switch quantity signals by the intelligent interface device (1) and then sent to the unmodified interval circuit breaker protection devices and the operation box (4).

3. A transition method for intelligent transformation of substation protection equipment according to claim 2, characterized in that: The intelligent interface device (1) comprises a PLC controller and a photoelectric conversion module. The PLC controller is used to receive switch signals from an unmodified interval circuit breaker protection device and an operation box (4), and convert the signals into optical digital signals through the photoelectric conversion module and send them to the intelligent bus protection device (2). At the same time, the PLC controller is also used to receive optical digital signals emitted by the intelligent bus protection device (2), and convert the signals into switch signals through the photoelectric conversion module and send them to the unmodified interval circuit breaker protection device and the operation box (4).

4. A transition method for intelligent transformation of substation protection equipment according to claim 3, characterized in that: The PLC controller is also used to perform status analysis on the received switch signal. When an abnormal state is detected, the PLC controller can trigger emergency response measures, including issuing an alarm signal or automatically cutting off related circuits.

5. The transition method for intelligent transformation of substation protection equipment according to claim 1, characterized in that: The intelligent interface device (1) has a built-in processor and a storage module, and an integrated sensor module. The processor and the storage module are used for the built-in self-learning artificial intelligence algorithm and the storage of historical data. The sensor module is used for real-time monitoring of signal transmission delay, noise, voltage fluctuation and other parameters. The intelligent interface device (1) can automatically adjust the signal conversion parameters according to the operating status of the substation, optimize the signal transmission efficiency, and form a signal transmission optimization model based on the self-learning artificial intelligence algorithm. The specific process is as follows: The long short-term memory network (LSTM) is used as the core algorithm to model and predict parameters such as delay and noise in signal transmission; Establishment of signal transmission optimization model: 1) Data collection and preprocessing: Collect historical signal transmission data, including: Signal delay D(t)D(t), noise level N(t)N(t), voltage fluctuation V(t)V(t) and signal packet loss rate L(t)L(t); Normalize the data: Among them, X is the original data, Xmin and Xmax are the minimum and maximum values ​​of the data respectively; 2) Model training Input data: time series data {D(t), N(t), V(t), L(t)}{D(t), N(t), V(t), L(t)}; Output data: optimized signal transmission parameters {P(t)}{P(t)}, such as sampling rate and filtering strength; 3) Model optimization: Use the Adam optimizer to train the model and set the learning rate. During the training process, use the early stopping method to prevent overfitting. 4) Model deployment: Deploy the trained LSTM model to the intelligent interface device to collect signal transmission data in real time and generate optimization strategies.

6. A transition method for intelligent transformation of substation protection equipment according to claim 5, characterized in that: The intelligent interface device (1) is trained to automatically select the optimal signal processing strategy under different operating environments through a reinforcement learning algorithm. The specific process is as follows: The deep Q network DQN is used as the core algorithm to train the intelligent interface device to automatically select the optimal signal processing strategy under different operating environments; Establishment of reinforcement learning algorithm: 1) Define the state space, action space and reward function: State space S: includes signal delay D(t)D(t), noise level N(t)N(t), voltage fluctuation V(t)V(t) and signal packet loss rate L(t)L(t); Action space A: adjust sampling rate and filter strength; Reward function R: R(t)=-αD(t)-βN(t)-γL(t); Among them, α, β, and γ are weight coefficients used to balance the impact of delay, noise, and packet loss rate; 2) Building a deep Q network: Input layer: the input dimension is 4, corresponding to 4 state features); Hidden layer: contains two fully connected layers, each with 128 neurons; Output layer: The output dimension is 4, corresponding to 4 actions; Loss function: Mean square error MSE is used as the loss function: Among them, Q pred (s,a) is the predicted Q value, Q target (s,a) is the target Q value; 3) Training: Experience replay: store historical states, actions, rewards, and next states for training; Target network: Use the target network to calculate the target Q value to reduce fluctuations during training; Q value update formula: Q target (s,a)=R(t)+γmax a′ Q(s′,a′); Among them, γ is the discount factor, s′ is the next state, and a′ is the next action; 4) Strategy selection: Use the ε-greedy strategy to select actions: Among them, ∈ is the exploration rate, the initial value is 1, and it gradually decays; 5) Model deployment: Deploy the trained DQN model to the intelligent interface device to collect state data in real time and select the optimal action.

7. A transition method for intelligent transformation of substation protection equipment according to claim 1, characterized in that: The intelligent interface device (1) adopts multi-channel parallel signal processing technology, supports simultaneous processing of multiple interval switch quantity signals, and improves signal processing speed and system response capability.

8. A transition method for intelligent transformation of substation protection equipment according to claim 7, characterized in that: The multi-channel parallel signal processing technology is implemented by a multi-core processor or FPGA.

9. A transition method for intelligent transformation of substation protection equipment according to claim 1, characterized in that: The intelligent interface device (1) has a fault self-healing function and can automatically recover or switch to a backup module when a fault is detected. The fault self-healing function is achieved through redundant design and a fault detection algorithm, ensuring that the device can automatically switch to a backup module when a fault occurs.

10. A transition method for intelligent transformation of substation protection equipment according to claim 9, characterized in that: The intelligent interface device (1) also automatically repairs software or configuration errors through a self-healing algorithm to reduce manual intervention.

Images