Remote control and data transmission method for circuit breaker switch based on sensor network

Abstract

This invention discloses a remote control and data transmission method for circuit breakers based on sensor networks, comprising: Step 1: collecting electrical parameter data, mechanical motion parameter data, thermal parameter data, and arc parameter data during the circuit breaker's operation to generate multi-source motion sensor data; Step 2: generating an action phase sequence based on the multi-source motion sensor data; Step 3: extracting the corresponding features of each action phase and generating an action causal token sequence; Step 4: inputting the action causal token sequence into an improved LightESD model to generate an action chain drift field and causal chain breakage results; Step 5: generating a remote control license code; Step 6: executing remote control of the circuit breaker and generating action execution results; Step 7: generating a causal compressed data packet and uploading it. This invention achieves circuit breaker action anomaly identification and remote control verification.

Description

Technical Field

[0001] This invention relates to the fields of artificial intelligence and intelligent control technology for power equipment, and in particular to a method for remote control and data transmission of circuit breaker switches based on sensor networks. Background Technology

[0002] With the rapid development of smart grids, distribution automation, and edge sensing technologies, circuit breakers, as core control devices in power systems, undertake critical tasks such as line disconnection, fault isolation, and remote control. For remote control and operational status monitoring of circuit breakers, existing technologies mainly employ remote communication modules combined with status acquisition devices to achieve circuit breaker action control and status data uploading. However, in actual operating scenarios, the following problems commonly exist:

[0003] Traditional remote control methods for circuit breakers mostly rely on single state parameters such as current, voltage, or switch position to determine the action. This makes it difficult to comprehensively capture the dynamic correlations between electrical changes, mechanical motion changes, thermal changes, and arc changes during circuit breaker operation. Consequently, the ability to identify complex abnormal states such as mechanical jamming, arcing anomalies, and action delays is insufficient. Existing condition monitoring methods typically use static thresholds or ordinary time-series analysis to determine the circuit breaker's operating state, lacking the ability to dynamically model the continuous causal relationships of the operation process. When circuit breakers experience mechanical wear, thermal decay, and changes in arc-extinguishing performance over long-term operation, existing methods struggle to adapt to dynamic changes. Adjusting control strategies in real time based on the evolving state can easily lead to remote miscontrol or missed detection of abnormal actions. Most existing edge intelligence detection methods only perform anomaly identification for single actions, without establishing a dynamic competition analysis mechanism between the real action chain and the historical action chain, nor forming a correlation modeling structure between the action drift field and causal break, making it difficult to achieve continuous evolution analysis during the circuit breaker operation process. In addition, existing remote data upload methods usually upload the raw monitoring data as a whole, lacking a key segment extraction mechanism based on the causal relationship of actions, resulting in high communication load and making it difficult to meet the real-time data transmission requirements of low-latency remote control scenarios at the edge.

[0004] Therefore, how to provide a remote control and data transmission method for circuit breaker switches based on sensor networks is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] One objective of this invention is to propose a remote control and data transmission method for circuit breakers based on sensor networks. This invention constructs a multi-source motion sensing structure that integrates electrical parameters, mechanical motion parameters, thermal parameters, and arc parameters. It combines an improved LightESD model with a motion causal chain analysis mechanism, utilizes the competitive offset relationship between the predicted motion mirror chain and the actual motion chain to generate a motion chain drift field, and dynamically corrects the predicted motion mirror chain based on the drift accumulation state. This enables circuit breaker motion anomaly identification, remote control verification, and causal compressed data transmission.

[0006] The remote control and data transmission method for circuit breakers based on sensor networks according to an embodiment of the present invention includes the following steps:

[0007] Step 1: Collect electrical parameter data, mechanical motion parameter data, thermal parameter data, and arc parameter data during the circuit breaker's opening and closing actions to generate multi-source motion sensor data;

[0008] Step 2: Based on the multi-source motion sensor data, perform phase division on the circuit breaker switch operation process to generate an action phase sequence;

[0009] Step 3: Based on the action phase sequence, extract the trigger time, duration, phase energy value, phase transition direction, and sensor source identifier corresponding to each action phase to generate an action causal token sequence;

[0010] Step 4: Input the action causal token sequence into the improved LightESD model, generate an action chain drift field by predicting the competitive matching between the action mirror chain and the real action chain, and generate a causal chain break result based on the action chain drift field;

[0011] Step 5: Based on the causal chain breakage result and the action chain drift field, generate the corresponding remote control license code;

[0012] Step Six: Execute the corresponding opening or closing action based on the remote control authorization code, collect action feedback data during the execution process, and generate action execution results;

[0013] Step 7: Based on the action chain drift field, causal chain breakage result, remote control license code, and action execution result, generate the corresponding causal compressed data packet, and upload the causal compressed data packet according to the action risk level.

[0014] Optionally, step one specifically includes:

[0015] Current acquisition nodes are installed at the main circuit conductors of the circuit breaker, voltage acquisition nodes are installed at the incoming and outgoing ends of the circuit breaker, displacement acquisition nodes and vibration acquisition nodes are installed at the operating mechanism of the circuit breaker, temperature acquisition nodes are installed at the contact points of the circuit breaker, and arc acquisition nodes are installed at the arc extinguishing chamber.

[0016] The system collects data on main circuit current changes, main circuit voltage changes, mechanism displacement changes, mechanism vibration changes, contact temperature changes, and arc intensity changes during the circuit breaker's opening and closing operations using each data acquisition node according to a unified sampling period, and generates corresponding sampling time identifiers based on a unified clock source.

[0017] Based on the sampling time identifiers corresponding to each acquisition node, time sorting and time alignment processing are performed on the main circuit current change data, main circuit voltage change data, mechanism displacement change data, mechanism vibration change data, contact temperature change data, and arc intensity change data to form a multi-channel synchronous action data stream corresponding to the sampling time.

[0018] Based on multi-channel synchronous action data stream, the changes in current, voltage, displacement, vibration, temperature and arc intensity between adjacent sampling times are calculated, and action change identifiers are generated according to the increase or decrease relationship of each change.

[0019] Based on the continuous change intervals corresponding to the action change identifiers, the multi-channel synchronous action data stream is divided into action stages to form an action stage index corresponding to each sampling time.

[0020] Based on the action stage index, stage correlation processing is performed on the main circuit current change data, main circuit voltage change data, mechanism displacement change data, mechanism vibration change data, contact temperature change data, and arc intensity change data to form multi-source action sensor data.

[0021] Optionally, step two specifically involves:

[0022] Read the main circuit current change data, main circuit voltage change data, mechanism displacement change data, mechanism vibration change data, contact temperature change data, and arc intensity change data from the multi-source motion sensor data, and arrange them in the order of execution time according to the sampling time marker;

[0023] Based on the changes in current, voltage, displacement, vibration, temperature, and arc intensity between adjacent sampling times, the motion change trajectory at the corresponding sampling time is constructed.

[0024] Based on the continuous change relationship between the various changes in the motion trajectory, the main circuit current start interval, mechanism displacement change interval, mechanism displacement termination interval, arc intensity change interval, and main circuit voltage recovery interval are identified.

[0025] Based on the time connection relationship corresponding to the main circuit current start interval, mechanism displacement change interval, mechanism displacement termination interval, arc intensity change interval and main circuit voltage recovery interval, the coil excitation phase, mechanism release phase, contact movement phase, arc formation phase, arc extinguishing phase and voltage recovery phase are divided.

[0026] Based on the start sampling time and end sampling time corresponding to each phase, the phase time boundary corresponding to each phase is generated.

[0027] According to the time sequence corresponding to the phase time boundary, the phase arrangement process is performed on the coil excitation phase, mechanism release phase, contact movement phase, arc formation phase, arc extinguishing phase and voltage recovery phase to generate the action phase sequence.

[0028] Optionally, step three specifically includes:

[0029] Read the start sampling time and end sampling time corresponding to each action phase in the action phase sequence, and generate the trigger time of the corresponding action phase based on the start sampling time;

[0030] Based on the start and end sampling times of each action phase, the time span of the corresponding action phase is calculated, and the duration of the corresponding action phase is generated.

[0031] Read the main circuit current change data, main circuit voltage change data, mechanism displacement change data, mechanism vibration change data, contact temperature change data, and arc intensity change data within the time interval corresponding to each action phase, and perform cumulative calculation on the corresponding change amount and sampling time interval at each sampling moment to generate the phase energy value of the corresponding action phase.

[0032] Based on the increase or decrease relationship of the change between consecutive sampling moments within the time interval corresponding to each action phase, the direction of continuous increase or decrease of the change corresponding to each action phase is identified, and the phase transition direction of the corresponding action phase is generated.

[0033] Based on the acquisition node number that participated in the calculation of the change within the time interval corresponding to each action phase, a sensor source identifier for the corresponding action phase is generated.

[0034] Based on the trigger time, duration, phase energy value, phase transition direction and sensor source identifier, phase feature association processing is performed on each action phase to form an action causal token for the corresponding action phase.

[0035] According to the time sequence corresponding to the action phase sequence, the action causal tokens corresponding to each action phase are processed by token arrangement to generate the action causal token sequence.

[0036] Optionally, step four specifically involves:

[0037] Input the action causal token sequence into the improved LightESD model;

[0038] The improved LightESD model includes an action mirror twin layer, a causal topological constraint layer, an action competition verification layer, an action chain drift field generation layer, a dynamic mirror feedback correction layer, and a causal break prediction layer.

[0039] The action mirror twin layer reads the historical action causal token sequence corresponding to the current action type, extracts historical action tokens according to the action phase, and generates mirror action tokens based on the trigger time, duration, phase energy value, phase transition direction and sensor source identifier corresponding to the historical action tokens. The mirror action tokens are arranged in the action phase time order to generate the predicted action mirror chain.

[0040] The causal topology constraint layer establishes the phase connection relationship between the coil excitation phase, the mechanism release phase, the contact movement phase, the arc formation phase, the arc extinguishing phase, and the voltage recovery phase. Based on the triggering time sequence, duration change relationship, and phase transition direction change relationship between adjacent action causal tokens, it identifies the phase missing state, phase lag state, phase advance state, and phase reversal state in the action causal token sequence.

[0041] The action competition verification layer establishes token matching pairs between real action tokens and mirror action tokens according to phase type and time sequence, and generates competition offsets based on trigger time difference, duration difference, phase energy value difference, and phase transition direction difference. For action tokens that have not established token matching pairs, an isolated token identifier is generated, and each competition offset is arranged in time sequence to generate the action chain competition offset trajectory.

[0042] The motion chain drift field generation layer performs drift aggregation processing on each competing offset in the motion chain competitive offset trajectory to generate time drift, energy drift, thermal drift, arc drift and mechanism inertial drift, and constructs the motion chain drift field.

[0043] The dynamic mirror feedback correction layer calculates the time correction, duration correction, and phase energy correction of the corresponding mirror action token in the predicted action mirror chain based on the drift direction and drift accumulation state of each drift amount, updates the corresponding mirror action token based on each correction amount, regenerates the predicted action mirror chain, and feeds the regenerated predicted action mirror chain back to the action competition verification layer to perform the competition matching of the next action cycle.

[0044] The causal break prediction layer retrieves the continuous drift intervals corresponding to each drift amount in the action chain drift field, and generates causal chain break results including the break phase name, break type, and break location based on the action phase, drift direction, and drift accumulation state corresponding to the continuous drift intervals.

[0045] Optionally, step five specifically includes:

[0046] Read the fracture phase name, fracture type and fracture location from the causal chain fracture results, and read the time drift, energy drift, thermal drift, arc drift and mechanism inertial drift from the action chain drift field;

[0047] Based on the phase connection relationship between the fracture phase name and the corresponding action phase, a fracture association chain is established between the corresponding action phases; based on the drift direction, drift duration and drift accumulation state corresponding to the time drift, energy drift, thermal drift, arc drift and mechanism inertial drift, the action risk value corresponding to each fracture association chain is calculated.

[0048] Risk levels are determined according to the numerical range corresponding to the risk values ​​of each action, and risk level labels for the corresponding action phases are generated.

[0049] Based on the fracture type, fracture location, and risk level identifier, a mapping relationship between action risk value and remote control status is established;

[0050] Based on the mapping relationship between action risk value and remote control status, status matching processing is performed on the control status corresponding to opening action, closing action and reset action to generate remote control permission status for the corresponding action type.

[0051] Generate the corresponding remote control license code according to the action type and remote control license status.

[0052] Optionally, step six specifically includes:

[0053] Read the action type identifier, control status identifier, and risk level identifier from the remote control license code, and parse the action direction, action duration, and action response window of the corresponding action type;

[0054] Based on the action type identifier, the action direction, action duration, and action response window, send the corresponding opening control command, closing control command, or reset control command to the circuit breaker operating mechanism.

[0055] During the execution of corresponding control commands by the circuit breaker operating mechanism, the main circuit current change data, main circuit voltage change data, mechanism displacement change data, mechanism vibration change data, contact temperature change data, and arc intensity change data are continuously collected in the order of sampling time to form an action feedback data stream.

[0056] Based on the continuous change relationship between the corresponding changes at each sampling moment in the action feedback data stream, the action start moment, mechanism response moment, contact action moment, arc change moment, and voltage recovery moment are identified.

[0057] Establish a time connection relationship according to the time sequence of action start time, mechanism response time, contact action time, arc change time and voltage recovery time, and generate the action feedback chain of the corresponding control action.

[0058] Based on the time interval, change of quantity, action duration and state switching order of each action in the action feedback chain, the action completion state, action delay state, action interruption state and action abnormal state are identified during the action execution process.

[0059] Based on the action completion state, action delay state, action interruption state, and action abnormal state, the corresponding control action execution result is generated.

[0060] Optionally, step seven specifically includes:

[0061] Read the time drift, energy drift, thermal drift, arc drift and mechanism inertial drift in the action chain drift field; read the fracture phase name, fracture type and fracture location in the causal chain fracture result; read the action type identifier, control status identifier and risk level identifier in the remote control license code; and read the action completion status, action delay status, action interruption status and action abnormal status in the action execution result.

[0062] Based on the drift direction, drift duration, and drift accumulation state corresponding to the time drift, energy drift, thermal drift, arc drift, and mechanism inertial drift, motion drift related segments are established according to the time sequence corresponding to the motion phase.

[0063] Based on the fracture phase name, fracture type, and fracture location, establish the phase connection relationship between the fracture phase and the motion drift related segment, and generate causal related segments.

[0064] Based on the action type identifier, control status identifier, and risk level identifier, control status association processing is performed on the causal association fragment to generate a control association fragment.

[0065] Based on the action completion state, action delay state, action interruption state, and action abnormal state, the action result association processing is performed on the control-related segments in the order of action execution time to generate an action causal association chain.

[0066] Extract continuous drift intervals and broken correlation intervals from the action causal relationship chain, and perform compression encoding on the action drift correlation segment, causal correlation segment, control correlation segment, and action causal relationship chain according to the corresponding time sequence of the action causal relationship chain to generate causal compressed data packets;

[0067] Based on the risk level of the action corresponding to the risk level identifier, determine the upload priority, upload data length, upload time interval, and upload frequency of the corresponding causal compressed data packet, and execute the upload of the causal compressed data packet according to the upload priority, upload data length, upload time interval, and upload frequency.

[0068] The beneficial effects of this invention are:

[0069] This invention addresses the issues of insufficient correlation of action states, difficulty in modeling continuous action evolution, and low accuracy in identifying abnormal actions during remote control of circuit breakers by constructing a multi-source action perception structure that integrates electrical parameters, mechanical motion parameters, thermal parameters, and arc parameters. It combines an improved LightESD model with a collaborative analysis mechanism of action causal chains. The invention proposes an action chain modeling strategy based on action phase partitioning, action causal token construction, and the generation of predicted action mirror chains, significantly improving the continuous state representation and dynamic anomaly perception capabilities of complex circuit breaker actions. Furthermore, the model structure introduces an action mirror twin layer, a causal topological constraint layer, and an action competition verification layer. By constructing the action chain competition offset trajectory through the competitive offset relationship between the actual action chain and the predicted action mirror chain, it achieves the detection of phase missing states and phases during circuit breaker operation. Dynamic identification of hysteresis states and phase reversal states; in the drift analysis stage, a motion chain drift field generation layer and a dynamic mirror feedback correction layer are constructed. Combined with the drift accumulation state in the continuous action cycle, the predicted motion mirror chain is dynamically iteratively updated, so that the predicted motion mirror chain evolves synchronously with the wear state of the mechanism, the arc decay state, and the thermal change state, effectively enhancing the model's adaptability to changes in the long-term operating state of the circuit breaker; in the control stage, a mapping relationship between the fracture correlation chain and the remote control state is established, and a remote control permission code is generated based on the action risk value to realize dynamic control verification during the remote action process of the circuit breaker; finally, using the causal compressed data packet generation mechanism, key causal segments are extracted and compressed and encoded in the continuous drift interval and fracture correlation interval to realize causal correlation compressed transmission and low-latency edge uploading of circuit breaker action data. Attached Figure Description

[0070] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0071] Figure 1 This is a schematic diagram of the overall process of the remote control and data transmission method for circuit breaker switches based on sensor networks proposed in this invention;

[0072] Figure 2 This is a schematic diagram of the improved LightESD model in the remote control and data transmission method for circuit breakers based on sensor networks proposed in this invention. Detailed Implementation

[0073] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0074] refer to Figures 1-2 A method for remote control and data transmission of circuit breaker switches based on sensor networks includes the following steps:

[0075] Step 1: Collect electrical parameter data, mechanical motion parameter data, thermal parameter data, and arc parameter data during the circuit breaker's opening and closing actions to generate multi-source motion sensor data;

[0076] Step 2: Based on multi-source motion sensor data, perform phase division on the circuit breaker switch operation process and generate an action phase sequence;

[0077] Step 3: Based on the action phase sequence, extract the trigger time, duration, phase energy value, phase transition direction, and sensor source identifier corresponding to each action phase to generate an action causal token sequence;

[0078] Step 4: Input the action causal token sequence into the improved LightESD model, generate the action chain drift field by predicting the competitive matching between the action mirror chain and the real action chain, and generate the causal chain break result based on the action chain drift field;

[0079] Step 5: Based on the causal chain breakage results and the action chain drift field, generate the corresponding remote control license code;

[0080] Step Six: Execute the corresponding opening or closing action based on the remote control license code, collect action feedback data during the execution process, and generate action execution results;

[0081] Step 7: Based on the action chain drift field, causal chain breakage results, remote control license code, and action execution results, generate the corresponding causal compressed data packet and upload the causal compressed data packet according to the action risk level.

[0082] In this embodiment, step one specifically includes:

[0083] Current acquisition nodes are installed at the main circuit conductors of the circuit breaker, voltage acquisition nodes are installed at the incoming and outgoing ends of the circuit breaker, displacement acquisition nodes and vibration acquisition nodes are installed at the operating mechanism of the circuit breaker, temperature acquisition nodes are installed at the contact points of the circuit breaker, and arc acquisition nodes are installed at the arc extinguishing chamber.

[0084] The system collects data on main circuit current changes, main circuit voltage changes, mechanism displacement changes, mechanism vibration changes, contact temperature changes, and arc intensity changes during the circuit breaker's opening and closing operations using each data acquisition node according to a unified sampling period, and generates corresponding sampling time identifiers based on a unified clock source.

[0085] Based on the sampling time identifiers corresponding to each acquisition node, time sorting and time alignment processing are performed on the main circuit current change data, main circuit voltage change data, mechanism displacement change data, mechanism vibration change data, contact temperature change data, and arc intensity change data to form a multi-channel synchronous action data stream corresponding to the sampling time.

[0086] Based on multi-channel synchronous action data stream, the changes in current, voltage, displacement, vibration, temperature and arc intensity between adjacent sampling times are calculated, and action change identifiers are generated according to the increase or decrease relationship of each change.

[0087] Based on the continuous change intervals corresponding to the action change identifiers, the multi-channel synchronous action data stream is divided into action stages to form an action stage index corresponding to each sampling time.

[0088] Based on the action stage index, stage correlation processing is performed on the main circuit current change data, main circuit voltage change data, mechanism displacement change data, mechanism vibration change data, contact temperature change data, and arc intensity change data to form multi-source action sensor data.

[0089] In this implementation, current acquisition nodes, voltage acquisition nodes, displacement acquisition nodes, vibration acquisition nodes, temperature acquisition nodes, and arc acquisition nodes are all connected to the internal clock bus of the edge control module. Each acquisition node performs synchronous sampling based on a unified clock pulse. The edge control module performs timing reconstruction on the data corresponding to different acquisition nodes based on the sampling time identifier, so that electrical changes, mechanical motion changes, thermal changes, and arc changes within the same action phase form a unified action association segment. During the action phase division process, the edge control module identifies the action phase boundary based on the continuous increase or decrease of the change between adjacent sampling times, and performs aggregation processing on the continuous change intervals within the action phase, so that the action phase index maintains a correspondence with the actual action process of the circuit breaker. During the phase association processing, the edge control module establishes the phase coupling relationship between different types of sensor data based on the action phase index, so that the electrical parameter changes, mechanical motion changes, thermal changes, and arc changes corresponding to the same action phase form a unified action semantic data structure.

[0090] In this embodiment, step two specifically involves:

[0091] Read the main circuit current change data, main circuit voltage change data, mechanism displacement change data, mechanism vibration change data, contact temperature change data, and arc intensity change data from the multi-source motion sensor data, and arrange them in the order of execution time according to the sampling time marker;

[0092] Based on the changes in current, voltage, displacement, vibration, temperature, and arc intensity between adjacent sampling times, the motion change trajectory at the corresponding sampling time is constructed.

[0093] Based on the continuous change relationship between the various changes in the motion trajectory, the main circuit current start interval, mechanism displacement change interval, mechanism displacement termination interval, arc intensity change interval, and main circuit voltage recovery interval are identified.

[0094] Based on the time connection relationship corresponding to the main circuit current start interval, mechanism displacement change interval, mechanism displacement termination interval, arc intensity change interval and main circuit voltage recovery interval, the coil excitation phase, mechanism release phase, contact movement phase, arc formation phase, arc extinguishing phase and voltage recovery phase are divided.

[0095] Based on the start sampling time and end sampling time corresponding to each phase, the phase time boundary corresponding to each phase is generated.

[0096] According to the time sequence corresponding to the phase time boundary, the phase arrangement process is performed on the coil excitation phase, mechanism release phase, contact movement phase, arc formation phase, arc extinguishing phase and voltage recovery phase to generate the action phase sequence.

[0097] In this implementation, the motion change trajectory is constructed using the correlation between changes in multi-channel synchronous motion data streams at continuous sampling times. The edge control module identifies the coil excitation process and the mechanism release process based on the time coupling relationship between the change in main circuit current and the change in mechanism displacement; it identifies the contact movement process and the arc formation process based on the continuous change relationship between the change in mechanism displacement and the change in arc intensity; and it identifies the arc extinguishing process and the voltage recovery process based on the attenuation relationship between the change in arc intensity and the change in main circuit voltage. During phase division, the temporal correlation between adjacent motion intervals is established by connecting them with continuous sampling times, so that a continuous motion chain structure is formed between each motion phase. During the generation of phase time boundaries, the edge control module determines the phase start boundary based on the sampling time when the change in the motion change trajectory first undergoes a continuous change, and determines the phase termination boundary based on the sampling time when the change in the change returns to a stable state.

[0098] In this embodiment, step three specifically includes:

[0099] Read the start sampling time and end sampling time corresponding to each action phase in the action phase sequence, and generate the trigger time of the corresponding action phase based on the start sampling time;

[0100] Based on the start and end sampling times of each action phase, the time span of the corresponding action phase is calculated, and the duration of the corresponding action phase is generated.

[0101] Read the main circuit current change data, main circuit voltage change data, mechanism displacement change data, mechanism vibration change data, contact temperature change data, and arc intensity change data within the time interval corresponding to each action phase, and perform cumulative calculation on the corresponding change amount and sampling time interval at each sampling moment to generate the phase energy value of the corresponding action phase.

[0102] Based on the increase or decrease relationship of the change between consecutive sampling moments within the time interval corresponding to each action phase, the direction of continuous increase or decrease of the change corresponding to each action phase is identified, and the phase transition direction of the corresponding action phase is generated.

[0103] Based on the acquisition node number that participated in the calculation of the change within the time interval corresponding to each action phase, a sensor source identifier for the corresponding action phase is generated.

[0104] Based on the trigger time, duration, phase energy value, phase transition direction and sensor source identifier, phase feature association processing is performed on each action phase to form an action causal token for the corresponding action phase.

[0105] According to the time sequence corresponding to the action phase sequence, the action causal tokens corresponding to each action phase are processed by token arrangement to generate the action causal token sequence.

[0106] In this implementation, the action causal token is constructed using a structured association form that corresponds to the multi-source change characteristics of the action phase. The edge control module generates a phase energy value based on the continuous cumulative relationship between the change amount at each sampling moment and the sampling time interval within the time interval corresponding to the action phase, so that the phase energy value reflects the intensity and duration of the change within the action phase. The phase transition direction is marked by the continuous increase or decrease of the change amount between consecutive sampling moments. A positive transition mark is generated when the change amount corresponding to multiple consecutive sampling moments remains in an increasing state, and a reverse transition mark is generated when the change amount corresponding to multiple consecutive sampling moments remains in a decreasing state. During the generation of the action causal token, the edge control module establishes the association relationship between the action phase and the sensor node based on the acquisition node number corresponding to the action phase, so that the action causal token simultaneously contains action time information, action change information, and action source information, and establishes an action causal association chain according to the time sequence corresponding to the action phase.

[0107] In this embodiment, step four specifically includes:

[0108] Input the action causal token sequence into the improved LightESD model;

[0109] The improved LightESD model includes an action mirror twin layer, a causal topological constraint layer, an action competition verification layer, an action chain drift field generation layer, a dynamic mirror feedback correction layer, and a causal break prediction layer.

[0110] The action mirror twin layer reads the historical action causal token sequence corresponding to the current action type, extracts historical action tokens according to the action phase, and generates mirror action tokens based on the trigger time, duration, phase energy value, phase transition direction and sensor source identifier corresponding to the historical action tokens. The mirror action tokens are arranged in the action phase time order to generate the predicted action mirror chain.

[0111] The causal topology constraint layer establishes the phase connection relationship between the coil excitation phase, mechanism release phase, contact movement phase, arc formation phase, arc extinguishing phase, and voltage recovery phase. Based on the triggering time sequence, duration change relationship, and phase transition direction change relationship between adjacent action causal tokens, it identifies the phase missing state, phase lag state, phase advance state, and phase reversal state in the action causal token sequence.

[0112] The action competition verification layer establishes token matching pairs between real action tokens and mirror action tokens according to phase type and time sequence, and generates competition offsets based on trigger time difference, duration difference, phase energy value difference, and phase transition direction difference. For action tokens that have not established token matching pairs, an isolated token identifier is generated, and each competition offset is arranged in time sequence to generate the action chain competition offset trajectory.

[0113] The motion chain drift field generation layer performs drift aggregation processing on each competing offset in the motion chain competitive offset trajectory to generate time drift, energy drift, thermal drift, arc drift and mechanism inertial drift, and constructs the motion chain drift field;

[0114] The dynamic mirror feedback correction layer calculates the time correction, duration correction, and phase energy correction of the corresponding mirror action token in the predicted action mirror chain based on the drift direction and drift accumulation state of each drift amount. It then updates the corresponding mirror action token based on each correction amount, regenerates the predicted action mirror chain, and feeds the regenerated predicted action mirror chain back to the action competition verification layer to perform the competition matching of the next action cycle.

[0115] The causal break prediction layer retrieves the continuous drift intervals corresponding to each drift amount in the action chain drift field, and generates causal chain break results including the break phase name, break type, and break location based on the action phase, drift direction, and drift accumulation state corresponding to the continuous drift intervals.

[0116] In this implementation, the predicted action mirror chain is dynamically constructed using historical action tokens corresponding to the current action type from the historical action causal token sequence. The action mirror twin layer establishes a mirror action parameter association table based on the time distribution relationship of historical action tokens in continuous action cycles. During the action competition verification process, the edge control module establishes a competition offset association matrix based on the phase correspondence between the real action token and the mirror action token, and generates the action chain competition offset trajectory based on the continuous change state of each competition offset in the competition offset association matrix. During the dynamic mirror feedback correction process, the edge control module iteratively updates the time parameters and energy parameters corresponding to the mirror action token based on the drift accumulation state in continuous action cycles, so that the predicted action mirror chain changes synchronously with the wear state, arc decay state, and thermal change state of the circuit breaker mechanism. During the causal break prediction process, the edge control module determines the action chain break position based on the continuous change relationship of the drift direction corresponding to the continuous drift interval.

[0117] Both the improved LightESD model and the LightESD model adopt a lightweight edge anomaly detection architecture. Both perform anomaly state recognition based on the temporal change relationship of the input sequence. Both include sequence input, state correlation analysis and anomaly result output processes. Both are suitable for low-computing-power real-time processing scenarios in edge control modules.

[0118] The improved LightESD model adds an action mirror twin layer, a causal topological constraint layer, an action competition verification layer, an action chain drift field generation layer, a dynamic mirror feedback correction layer, and a causal break prediction layer to the LightESD model. The action mirror twin layer generates a predicted action mirror chain based on historical action causal tokens. The action competition verification layer generates an action chain competition offset trajectory based on the competition offset relationship between the real action chain and the predicted action mirror chain. The dynamic mirror feedback correction layer performs dynamic updates on the predicted action mirror chain based on the drift accumulation state.

[0119] The improved LightESD model can dynamically adjust the predicted action mirror chain based on the drift changes during the continuous operation of the circuit breaker, so that the predicted action mirror chain keeps pace with the actual operation state of the circuit breaker. At the same time, the improved LightESD model uses the action chain competition offset trajectory to identify the phase missing state, phase reversal state and causal chain break position in the action chain, so that the remote control process of the circuit breaker has dynamic action verification capability and continuous action evolution analysis capability.

[0120] In this embodiment, step five specifically includes:

[0121] Read the fracture phase name, fracture type and fracture location from the causal chain fracture results, and read the time drift, energy drift, thermal drift, arc drift and mechanism inertial drift from the action chain drift field;

[0122] Based on the phase connection relationship between the fracture phase name and the corresponding action phase, a fracture association chain is established between the corresponding action phases; based on the drift direction, drift duration and drift accumulation state corresponding to the time drift, energy drift, thermal drift, arc drift and mechanism inertial drift, the action risk value corresponding to each fracture association chain is calculated.

[0123] Risk levels are determined according to the numerical range corresponding to the risk values ​​of each action, and risk level labels for the corresponding action phases are generated.

[0124] Based on the fracture type, fracture location, and risk level identifier, a mapping relationship between action risk value and remote control status is established;

[0125] Based on the mapping relationship between action risk value and remote control status, status matching processing is performed on the control status corresponding to opening action, closing action and reset action to generate remote control permission status for the corresponding action type.

[0126] Generate the corresponding remote control license code according to the action type and remote control license status.

[0127] In this implementation, the motion risk value is constructed using the continuous change relationship of the drift state corresponding to the fracture correlation chain. The edge control module establishes a multi-dimensional drift correlation matrix based on the cumulative change of time drift, energy drift, thermal drift, arc drift, and mechanism inertial drift in the continuous motion cycle, and generates the motion risk value based on the drift duration and drift direction change relationship of each drift quantity in the multi-dimensional drift correlation matrix. During the remote control state mapping process, the edge control module establishes the correlation between the motion risk value and the remote control state based on the connection position of the motion phase corresponding to the fracture position in the motion chain, so that the drift state corresponding to different motion phases is mapped to different remote control states. During the remote control license code generation process, the edge control module establishes a license coding rule based on the control state corresponding to the motion type and the risk level identifier corresponding to the motion risk value, and generates a remote control license code corresponding to the motion type.

[0128] In this embodiment, step six specifically includes:

[0129] Read the action type identifier, control status identifier, and risk level identifier from the remote control license code, and parse the action direction, action duration, and action response window of the corresponding action type;

[0130] Based on the action type identifier, the action direction, action duration, and action response window, send the corresponding opening control command, closing control command, or reset control command to the circuit breaker operating mechanism.

[0131] During the execution of corresponding control commands by the circuit breaker operating mechanism, the main circuit current change data, main circuit voltage change data, mechanism displacement change data, mechanism vibration change data, contact temperature change data, and arc intensity change data are continuously collected in the order of sampling time to form an action feedback data stream.

[0132] Based on the continuous change relationship between the corresponding changes at each sampling moment in the action feedback data stream, the action start moment, mechanism response moment, contact action moment, arc change moment, and voltage recovery moment are identified.

[0133] Establish a time connection relationship according to the time sequence of action start time, mechanism response time, contact action time, arc change time and voltage recovery time, and generate the action feedback chain of the corresponding control action.

[0134] Based on the time interval, change of quantity, action duration and state switching order of each action in the action feedback chain, the action completion state, action delay state, action interruption state and action abnormal state are identified during the action execution process.

[0135] Based on the action completion state, action delay state, action interruption state, and action abnormal state, the corresponding control action execution result is generated.

[0136] In this implementation, the action feedback chain is constructed using the continuous temporal connection relationship between each action moment during the action execution process. The edge control module identifies the mechanism response state based on the time interval between the action initiation moment and the mechanism response moment, identifies the contact execution state based on the time interval between the mechanism response moment and the contact action moment, and identifies the arc extinguishing recovery state based on the continuous change relationship between the arc change moment and the voltage recovery moment. During the action execution state identification process, the edge control module establishes a state switching association relationship based on the continuous increase or decrease of the change quantity corresponding to each action moment in the action feedback chain, and identifies the action completion state, action delay state, action interruption state, and action abnormal state based on the state switching association relationship. During the action execution result generation process, the edge control module establishes a mapping relationship of the action execution process based on the corresponding time sequence of the action feedback chain, so that the action execution result maintains a correspondence with the actual action process.

[0137] In this embodiment, step seven specifically includes:

[0138] Read the time drift, energy drift, thermal drift, arc drift and mechanism inertial drift in the action chain drift field; read the fracture phase name, fracture type and fracture location in the causal chain fracture result; read the action type identifier, control status identifier and risk level identifier in the remote control license code; and read the action completion status, action delay status, action interruption status and action abnormal status in the action execution result.

[0139] Based on the drift direction, drift duration, and drift accumulation state corresponding to the time drift, energy drift, thermal drift, arc drift, and mechanism inertial drift, motion drift related segments are established according to the time sequence corresponding to the motion phase.

[0140] Based on the fracture phase name, fracture type, and fracture location, establish the phase connection relationship between the fracture phase and the motion drift related segment, and generate causal related segments.

[0141] Based on the action type identifier, control status identifier, and risk level identifier, control status association processing is performed on the causal association fragment to generate a control association fragment.

[0142] Based on the action completion state, action delay state, action interruption state, and action abnormal state, the action result association processing is performed on the control-related segments in the order of action execution time to generate an action causal association chain.

[0143] Extract continuous drift intervals and broken correlation intervals from the action causal relationship chain, and perform compression encoding on the action drift correlation segment, causal correlation segment, control correlation segment, and action causal relationship chain according to the corresponding time sequence of the action causal relationship chain to generate causal compressed data packets;

[0144] Based on the risk level of the action corresponding to the risk level identifier, determine the upload priority, upload data length, upload time interval, and upload frequency of the corresponding causal compressed data packet, and execute the upload of the causal compressed data packet according to the upload priority, upload data length, upload time interval, and upload frequency.

[0145] In this implementation, the motion drift correlation segments are organized according to the time sequence of each drift quantity in the motion chain drift field. The edge control module establishes the motion drift correlation structure based on the time connection relationship between continuous drift intervals. During the generation of causal correlation segments, the edge control module establishes the phase correlation relationship between the fracture phase and the continuous drift interval based on the connection position of the fracture phase in the motion chain, so that the causal correlation segments simultaneously contain drift change information and fracture position information. During the construction of the motion causal correlation chain, the edge control module establishes the control state change relationship and the motion execution state change relationship based on the action execution time sequence, and extracts the key correlation intervals based on the time distribution state corresponding to the continuous drift intervals and fracture correlation intervals. During the generation of causal compressed data packets, the edge control module performs segmented compression encoding based on the time span and state switching density corresponding to the key correlation intervals, so that different action risk levels correspond to different data upload particles.

[0146] Example 1: To verify the feasibility of this invention in practice, it was applied to a remote control system for circuit breakers in a 110kV intelligent substation. The substation contains 12 high-voltage vacuum circuit breakers, deployed in incoming line cabinets, feeder cabinets, and bus tie cabinets. Each circuit breaker is equipped with an edge control module, current acquisition nodes, voltage acquisition nodes, displacement acquisition nodes, vibration acquisition nodes, temperature acquisition nodes, and arc acquisition nodes. The edge control module establishes a remote communication connection with the substation's main dispatch station via an industrial Ethernet network. During substation operation, due to the frequent opening and closing actions of the circuit breakers, problems such as mechanical wear, arc attenuation, and action delays are prone to occur. Traditional remote control systems typically rely solely on current status or switch position to determine actions, making it difficult to accurately identify the continuous changes between different action phases during circuit breaker operation. When mechanical jamming, arc abnormalities, or reverse action sequences occur, miscontrol and missed detection of abnormal actions are likely to occur. Furthermore, traditional upload methods typically upload the complete action waveform to the main station, resulting in high communication load and failing to meet the real-time upload requirements of high-frequency remote control scenarios.

[0147] In this embodiment, the edge control module first synchronously collects the main circuit current change data, main circuit voltage change data, mechanism displacement change data, mechanism vibration change data, contact temperature change data, and arc intensity change data during the circuit breaker's opening and closing actions. It then performs time alignment processing based on a unified clock bus to form a multi-channel synchronous action data stream. Subsequently, the edge control module divides the coil excitation phase, mechanism release phase, contact movement phase, arc formation phase, arc extinguishing phase, and voltage recovery phase according to the continuous change relationship between the changes in the amount of change at different sampling times, and further constructs an action causal token sequence.

[0148] After the action causal token sequence is input into the improved LightESD model, the action mirror twin layer generates a predicted action mirror chain based on the historical action causal token sequence. The action competition verification layer performs a competition matching analysis on the real action chain and the predicted action mirror chain. When the circuit breaker mechanism experiences slight wear, the duration of the mechanism release phase in the real action chain gradually increases, and the time drift in the action chain competition offset trajectory continues to increase. The dynamic mirror feedback correction layer performs time correction on the mirror action tokens in the predicted action mirror chain based on the drift accumulation state, so that the predicted action mirror chain keeps the actual action state of the circuit breaker in sync. When an abnormal arc attenuation occurs inside the arc extinguishing chamber, a continuous drift interval appears between the arc formation phase and the voltage recovery phase. The causal break prediction layer identifies the corresponding break location and generates the corresponding causal chain break result. The edge control module dynamically adjusts the remote control permission state based on the action risk value and restricts continuous closing operations for high-risk circuit breakers.

[0149] During the data upload process, the edge control module does not upload the complete action waveform. Instead, it extracts the continuous drift intervals and broken correlation intervals in the action causal chain, generates causal compressed data packets, and dynamically adjusts the upload frequency and upload length according to the action risk level. For low-risk actions, only key drift segments are uploaded; for high-risk actions, the upload length and upload frequency of broken correlation intervals are increased, thereby reducing the communication load on the edge side and improving the real-time performance of remote fault analysis.

[0150] To further verify the technical effect of the present invention, the method of the present invention was compared with the traditional threshold control method, the ordinary LSTM action recognition method and the ordinary edge anomaly detection method. The test period was 45 consecutive days, and a total of 12,684 sets of circuit breaker action data were collected, including 10,836 sets of normal action data, 1,042 sets of action delay data, 516 sets of arc anomaly data and 290 sets of mechanism jamming data. The experimental results are shown in Table 1.

[0151] Table 1. Comparison of test results for different methods in remote control scenarios of circuit breakers.

[0152] Traditional threshold control methods 81.6 73.4 68.2 27 148 5120 12.5 0.8 Ordinary LSTM action recognition method 88.9 84.1 79.5 16 126 4380 24.3 1.7 Common edge anomaly detection methods 91.3 87.8 83.6 11 118 3915 31.6 2.4 Method of the present invention 97.8 96.5 95.1 2 89 1286 74.8 6.9

[0153] As shown in Table 1, the method of this invention exhibits significantly better overall performance in remote control scenarios for circuit breakers than traditional threshold control methods, ordinary LSTM action recognition methods, and ordinary edge anomaly detection methods. The action anomaly recognition accuracy of this invention reaches 97.8%, a 16.2% improvement compared to traditional threshold control methods, indicating that the competitive matching mechanism based on action causal token sequences and predicted action mirror chains effectively enhances the ability to recognize complex action states. The action delay recognition rate and arc anomaly recognition rate reach 96.5% and 95.1%, respectively, demonstrating that this invention has a high perception capability for continuously evolving abnormal states such as mechanism wear and arc decay. Furthermore, the number of remote miscontrols using this invention is only 2, and the average action response time is reduced to 89ms, indicating that the dynamic mirror feedback correction mechanism can improve the stability and real-time performance of remote control. In addition, the data compression rate of this invention reaches 74.8%, significantly reducing the amount of data uploaded, verifying that the causal compressed data packet mechanism can effectively reduce the communication load on the edge side.

[0154] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for remote control and data transmission of circuit breaker switches based on sensor networks, characterized in that, Includes the following steps: Step 1: Collect electrical parameter data, mechanical motion parameter data, thermal parameter data, and arc parameter data during the circuit breaker's opening and closing actions to generate multi-source motion sensor data; Step 2: Based on the multi-source motion sensor data, perform phase division on the circuit breaker switch operation process to generate an action phase sequence; Step 3: Based on the action phase sequence, extract the trigger time, duration, phase energy value, phase transition direction, and sensor source identifier corresponding to each action phase to generate an action causal token sequence; Step 4: Input the action causal token sequence into the improved LightESD model, generate an action chain drift field by predicting the competitive matching between the action mirror chain and the real action chain, and generate a causal chain break result based on the action chain drift field; Step 5: Based on the causal chain breakage result and the action chain drift field, generate the corresponding remote control license code; Step 6: Execute the corresponding opening or closing action based on the remote control license code, collect action feedback data during the execution process, and generate action execution results; Step 7: Based on the action chain drift field, causal chain breakage result, remote control license code, and action execution result, generate the corresponding causal compressed data packet, and upload the causal compressed data packet according to the action risk level.

2. The sensor network based remote control and data transfer method for circuit breaker switch according to claim 1, characterized in that, Step one specifically involves: Current acquisition nodes are installed at the main circuit conductors of the circuit breaker, voltage acquisition nodes are installed at the incoming and outgoing ends of the circuit breaker, displacement acquisition nodes and vibration acquisition nodes are installed at the operating mechanism of the circuit breaker, temperature acquisition nodes are installed at the contact points of the circuit breaker, and arc acquisition nodes are installed at the arc extinguishing chamber. The system collects data on main circuit current changes, main circuit voltage changes, mechanism displacement changes, mechanism vibration changes, contact temperature changes, and arc intensity changes during the circuit breaker's opening and closing operations using each data acquisition node according to a unified sampling period, and generates corresponding sampling time identifiers based on a unified clock source. Based on the sampling time identifiers corresponding to each acquisition node, time sorting and time alignment processing are performed on the main circuit current change data, main circuit voltage change data, mechanism displacement change data, mechanism vibration change data, contact temperature change data, and arc intensity change data to form a multi-channel synchronous action data stream corresponding to the sampling time. Based on multi-channel synchronous action data stream, the changes in current, voltage, displacement, vibration, temperature and arc intensity between adjacent sampling times are calculated, and action change identifiers are generated according to the increase or decrease relationship of each change. Based on the continuous change intervals corresponding to the action change identifiers, the multi-channel synchronous action data stream is divided into action stages to form an action stage index corresponding to each sampling time. Based on the action stage index, stage correlation processing is performed on the main circuit current change data, main circuit voltage change data, mechanism displacement change data, mechanism vibration change data, contact temperature change data, and arc intensity change data to form multi-source action sensor data.

3. The sensor network based remote control and data transfer method for circuit breaker switches as claimed in claim 1 wherein, Step two specifically involves: Read the main circuit current change data, main circuit voltage change data, mechanism displacement change data, mechanism vibration change data, contact temperature change data, and arc intensity change data from the multi-source motion sensor data, and arrange them in the order of execution time according to the sampling time marker; Based on the changes in current, voltage, displacement, vibration, temperature, and arc intensity between adjacent sampling times, the motion change trajectory at the corresponding sampling time is constructed. Based on the continuous change relationship between the various changes in the motion trajectory, the main circuit current start interval, mechanism displacement change interval, mechanism displacement termination interval, arc intensity change interval, and main circuit voltage recovery interval are identified. Based on the time connection relationship corresponding to the main circuit current start interval, mechanism displacement change interval, mechanism displacement termination interval, arc intensity change interval and main circuit voltage recovery interval, the coil excitation phase, mechanism release phase, contact movement phase, arc formation phase, arc extinguishing phase and voltage recovery phase are divided. Based on the start sampling time and end sampling time corresponding to each phase, the phase time boundary corresponding to each phase is generated. According to the time sequence corresponding to the phase time boundary, the phase arrangement process is performed on the coil excitation phase, mechanism release phase, contact movement phase, arc formation phase, arc extinguishing phase and voltage recovery phase to generate the action phase sequence.

4. The sensor network based remote control and data transfer method for circuit breaker switches as claimed in claim 1, wherein, Step three specifically involves: Read the start sampling time and end sampling time corresponding to each action phase in the action phase sequence, and generate the trigger time of the corresponding action phase based on the start sampling time; Based on the start and end sampling times of each action phase, the time span of the corresponding action phase is calculated, and the duration of the corresponding action phase is generated. Read the main circuit current change data, main circuit voltage change data, mechanism displacement change data, mechanism vibration change data, contact temperature change data, and arc intensity change data within the time interval corresponding to each action phase, and perform cumulative calculation on the corresponding change amount and sampling time interval at each sampling moment to generate the phase energy value of the corresponding action phase. Based on the increase or decrease relationship of the change between consecutive sampling moments within the time interval corresponding to each action phase, the direction of continuous increase or decrease of the change corresponding to each action phase is identified, and the phase transition direction of the corresponding action phase is generated. Based on the acquisition node number that participates in the calculation of change within the time interval corresponding to each action phase, a sensor source identifier for the corresponding action phase is generated. Based on the trigger time, duration, phase energy value, phase transition direction and sensor source identifier, phase feature association processing is performed on each action phase to form an action causal token for the corresponding action phase. According to the time sequence corresponding to the action phase sequence, the action causal tokens corresponding to each action phase are processed to generate an action causal token sequence.

5. The sensor network based remote control and data transfer method for circuit breaker switches as claimed in claim 1 wherein, Step four specifically involves: Input the action causal token sequence into the improved LightESD model; The improved LightESD model includes an action mirror twin layer, a causal topological constraint layer, an action competition verification layer, an action chain drift field generation layer, a dynamic mirror feedback correction layer, and a causal break prediction layer. The action mirror twin layer reads the historical action causal token sequence corresponding to the current action type, extracts historical action tokens according to the action phase, and generates mirror action tokens based on the trigger time, duration, phase energy value, phase transition direction and sensor source identifier corresponding to the historical action tokens. The mirror action tokens are arranged in the action phase time order to generate the predicted action mirror chain. The causal topology constraint layer establishes the phase connection relationship between the coil excitation phase, the mechanism release phase, the contact movement phase, the arc formation phase, the arc extinguishing phase, and the voltage recovery phase. Based on the triggering time sequence, duration change relationship, and phase transition direction change relationship between adjacent action causal tokens, it identifies the phase missing state, phase lag state, phase advance state, and phase reversal state in the action causal token sequence. The action competition verification layer establishes token matching pairs between real action tokens and mirror action tokens according to phase type and time sequence, and generates competition offsets based on trigger time difference, duration difference, phase energy value difference, and phase transition direction difference. For action tokens that have not established token matching pairs, an isolated token identifier is generated, and each competition offset is arranged in time sequence to generate the action chain competition offset trajectory. The motion chain drift field generation layer performs drift aggregation processing on each competing offset in the motion chain competitive offset trajectory to generate time drift, energy drift, thermal drift, arc drift and mechanism inertial drift, and constructs the motion chain drift field. The dynamic mirror feedback correction layer calculates the time correction, duration correction, and phase energy correction of the corresponding mirror action token in the predicted action mirror chain based on the drift direction and drift accumulation state of each drift amount, updates the corresponding mirror action token based on each correction amount, regenerates the predicted action mirror chain, and feeds the regenerated predicted action mirror chain back to the action competition verification layer to perform the competition matching of the next action cycle. The causal break prediction layer retrieves the continuous drift intervals corresponding to each drift amount in the action chain drift field, and generates causal chain break results including the break phase name, break type, and break location based on the action phase, drift direction, and drift accumulation state corresponding to the continuous drift intervals.

6. The remote control and data transmission method for circuit breaker switches based on sensor networks according to claim 1, characterized in that, Step five specifically involves: Read the fracture phase name, fracture type and fracture location from the causal chain fracture results, and read the time drift, energy drift, thermal drift, arc drift and mechanism inertial drift from the action chain drift field; Based on the phase connection relationship between the fracture phase name and the corresponding action phase, a fracture association chain is established between the corresponding action phases; based on the drift direction, drift duration and drift accumulation state corresponding to the time drift, energy drift, thermal drift, arc drift and mechanism inertial drift, the action risk value corresponding to each fracture association chain is calculated. Risk levels are determined according to the numerical range corresponding to the risk values ​​of each action, and risk level labels for the corresponding action phases are generated. Based on the fracture type, fracture location, and risk level identifier, a mapping relationship between action risk value and remote control status is established; Based on the mapping relationship between action risk value and remote control status, status matching processing is performed on the control status corresponding to opening action, closing action and reset action to generate remote control permission status for the corresponding action type. Generate the corresponding remote control license code according to the action type and remote control license status.

7. The remote control and data transmission method for circuit breaker switches based on sensor networks according to claim 1, characterized in that, Step six specifically involves: Read the action type identifier, control status identifier, and risk level identifier from the remote control license code, and parse the action direction, action duration, and action response window of the corresponding action type; Based on the action type identifier, the action direction, action duration, and action response window, send the corresponding opening control command, closing control command, or reset control command to the circuit breaker operating mechanism. During the execution of corresponding control commands by the circuit breaker operating mechanism, the main circuit current change data, main circuit voltage change data, mechanism displacement change data, mechanism vibration change data, contact temperature change data, and arc intensity change data are continuously collected in the order of sampling time to form an action feedback data stream. Based on the continuous change relationship between the corresponding changes at each sampling moment in the action feedback data stream, the action start moment, mechanism response moment, contact action moment, arc change moment, and voltage recovery moment are identified. Establish a time connection relationship according to the time sequence of action start time, mechanism response time, contact action time, arc change time and voltage recovery time, and generate the action feedback chain of the corresponding control action. Based on the time interval, change of quantity, action duration and state switching order of each action in the action feedback chain, the action completion state, action delay state, action interruption state and action abnormal state are identified during the action execution process. Based on the action completion state, action delay state, action interruption state, and action abnormal state, the corresponding action execution result is generated.

8. The sensor network based remote control and data transfer method for circuit breaker switches as claimed in claim 1, wherein, Step seven specifically involves: Read the time drift, energy drift, thermal drift, arc drift and mechanism inertial drift in the action chain drift field; read the fracture phase name, fracture type and fracture location in the causal chain fracture result; read the action type identifier, control status identifier and risk level identifier in the remote control license code; and read the action completion status, action delay status, action interruption status and action abnormal status in the action execution result. Based on the drift direction, drift duration, and drift accumulation state corresponding to the time drift, energy drift, thermal drift, arc drift, and mechanism inertial drift, motion drift related segments are established according to the time sequence corresponding to the motion phase. Based on the fracture phase name, fracture type, and fracture location, establish the phase connection relationship between the fracture phase and the motion drift related segment, and generate causal related segments. Based on the action type identifier, control status identifier, and risk level identifier, control status association processing is performed on the causal association fragment to generate a control association fragment. Based on the action completion state, action delay state, action interruption state, and action abnormal state, the action result association processing is performed on the control-related segments in the order of action execution time to generate an action causal association chain. Extract continuous drift intervals and broken correlation intervals from the action causal relationship chain, and perform compression encoding on the action drift correlation segment, causal correlation segment, control correlation segment, and action causal relationship chain according to the corresponding time sequence of the action causal relationship chain to generate causal compressed data packets; Based on the risk level of the action corresponding to the risk level identifier, determine the upload priority, upload data length, upload time interval, and upload frequency of the corresponding causal compressed data packet, and execute the upload of the causal compressed data packet according to the upload priority, upload data length, upload time interval, and upload frequency.

Images