Transformer substation operation and maintenance management method and system based on intelligent voice control

Abstract

According to the intelligent voice control substation operation and maintenance management method and system provided by the invention, the system realizes remote monitoring, automatic routing inspection, intelligent alarm and safety control of the substation through cooperative work of the voice assistant module, the remote control module and other functional modules. According to the method, the security of voiceprint recognition verification is effectively improved, the capability of preventing false voiceprint attacks in a complex scene is enhanced, and more efficient and reliable technical support is provided for a remote intelligent patrol system.

Claims

1. An intelligent voice-controlled substation operation and maintenance management method, characterized in that: include: Step 1: receiving user voice input, performing voice recognition on the input voice, and transmitting the recognized voice command to the intelligent control module; Step 2: The intelligent control module receives and analyzes the voice command, determines the type of the voice command, and executes corresponding steps; Step 3: Convert the system response into speech output to the user.

2. The method of claim 1, wherein: The step 1, receiving voice input through the voice assistant module, performing voice recognition on the input voice, and transmitting the recognized voice command to the intelligent control module, specifically includes: Step 1-1, collecting user voice input through a microphone; Step 1-2, performing voice recognition on the voice input to form a text instruction; Step 1-3: passing the text instruction to the intelligent control module through the API interface.

3. The method of claim 1, wherein: Step 2: The intelligent control module receives and analyzes the voice command, determines the type of the voice command, and executes corresponding steps, specifically including: If it is an inspection task instruction, the inspection task will be executed and the inspection report will be returned; If it is an alarm information query instruction, the device status is monitored, and an alarm is triggered when the threshold is reached or an abnormality occurs. The alarm information is returned, and a solution is matched and pushed; If it is a dashboard data query instruction, the visualized key performance indicators and dashboard data are returned; If it is a remote control instruction, the voiceprint recognition verification is performed, and if the verification is passed, the remote control operation is performed, otherwise the remote control request is rejected.

4. The method according to claim 3, wherein if the remote control instruction is a remote control instruction, voiceprint recognition verification is performed, and if the verification is passed, the remote control operation is performed, otherwise the remote control request is rejected, specifically comprising: Step 2-1, performing feature extraction on the voice input; Step 2-2: Perform voiceprint comparison and pseudo voiceprint detection; Step 2-3, verification result processing; Step 2-4: Perform operation records and security feedback, record verification results and operation logs. If a pseudo-voiceprint attack is detected, the system will automatically trigger a security alarm, record the attack information in the database and push it to the operation and maintenance personnel.

5. The method of claim 4, wherein: Step 2-1, extracting features from the voice input, specifically comprising: Step 2-1-1, extracting the basic spectral features of the speech signal and generating a basic feature vector; Step 2-1-2: Extract micro-vibration characteristics, extract the local feature differences of the speech signal based on the dynamic changes of high-frequency and low-frequency characteristics, and calculate using the following formula: Among them, F high (t) and F low (t) are the high-frequency and low-frequency signal strengths, respectively. t1 and t2 are time intervals used to capture the dynamic change characteristics of voiceprint details. Step 2-1-3, obtaining multi-dimensional voiceprint features, wherein the multi-dimensional voiceprint features include basic features and micro-vibration features of the voiceprint.

6. The method of claim 4, wherein: Step 2-2: Perform voiceprint comparison and pseudo-voiceprint detection, including: Step 2-2-1: Perform a preliminary match between the extracted multi-dimensional voiceprint features and the pre-stored voiceprint database, and calculate the matching score S match , the specific calculation formula is: Among them, Input Feature is a multi-dimensional voiceprint feature vector, including basic feature vectors and micro-vibration feature vectors; Database Feature is the voiceprint template of the corresponding user in the database, S match is the matching score, ranging from [0,1]; Step 2-2-2: Perform fake voiceprint detection, generate adversarial network (GAN) model, and judge whether the input features have fake features, using the following formula: P fake =f GAN (Input Features) Among them, P fake is the probability of pseudo voiceprint, f GAN Classification function trained for adversarial networks; Step 2-2-3: Combine the matching score S match and the probability of pseudo voiceprint P fake , comprehensively evaluate the final verification confidence: S final =w1·S match -w2·P fake Among them, w1 and w2 are the weights of matching degree and pseudo voiceprint probability respectively.

7. The method of claim 4, wherein: Step 2-3: Verification result processing, including: According to the final verification confidence S final Perform identity authentication and operation permission control. If S final ≥T accept , pass the verification, and allow the user to perform remote control operations; if S final <T accept , reject the request and inform the user of the failure reason through the voice assistant, where T accept is the verification threshold.

8. The method of claim 6, wherein: In step 2-2-2, the classification function f trained by the adversarial network GAN , specifically: f GAN (Input Features) = σ(W3·ReLU(W2·tanh(W1·Input Features+b1)+b2)+b3) where W1, W2, W3 are the weight matrices of the corresponding layers, b1, b2, b3 are the bias terms of the corresponding layers; tanh is the hyperbolic tangent function of the first layer, ReLU is the rectified linear unit of the second layer, and σ is the sigmoid function of the output layer. Introduce time-frequency characteristic weight adjustment in the first layer weight matrix W1: W1[i,j]=α·F freq [i]+(1-α)·T time [j] Among them, F freq [i] is the frequency characteristic weight of the i-th dimension, T time [j] is the j-th dimension time characteristic weight, and α is the dynamic adjustment coefficient, which is dynamically updated based on at least one of the environmental noise and the input feature quality.

9. An intelligent voice-controlled substation operation and maintenance management system, used to execute the substation intelligent operation and maintenance management method according to any one of claims 1 to 8, characterized in that: Specifically includes the following modules: A voice assistant module is used to receive user voice input through the voice assistant module, perform voice recognition on the input voice, and transmit the recognized voice command to the intelligent control module; An intelligent control module is used to receive and analyze the voice command, determine the type of the voice command, and start the corresponding function module; if it is an inspection task command, the intelligent inspection module is started; if it is an alarm information query command, the intelligent alarm module is started; if it is a dashboard data query command, the dashboard module is started; if it is a remote control command, the remote control module is started; Intelligent inspection module, used to execute inspection tasks and return inspection reports, and execute response output module after completion; Intelligent alarm module, which is used to monitor the status of equipment, trigger alarms when thresholds are reached or abnormalities occur, return alarm information, match and push solutions, and execute response output modules after completion; The dashboard module is used to return visual key performance indicators and dashboard data, and execute the response output module after completion; A remote control module is used to perform voiceprint recognition verification. If the verification is passed, the remote control operation is performed, otherwise the remote control request is rejected and the response output module is executed after completion; The response output module is used to convert the system response into voice output to the user.

10. The system according to claim 9, wherein the remote control module specifically comprises: A feature extraction unit, used for extracting features from the speech input; A comparison and detection unit, used for voiceprint comparison and pseudo-voiceprint detection; A verification processing unit, used for verification result processing; The recording and feedback unit is used to perform operation recording and security feedback, record verification results and operation logs. If a pseudo-voiceprint attack is detected, the system will automatically trigger a security alarm, record the attack information in the database and push it to the operation and maintenance personnel.

Images