A power equipment safety adjusting system based on data analysis

Abstract

The application belongs to the technical field of power system automatic control, and relates to a power equipment safety regulation system based on data analysis, which comprises a data acquisition unit, a state space analysis unit and a nonlinear adaptive regulation unit; a running state space is constructed by using a real-time output state sequence, a state offset modulus of a state point relative to a safety boundary is solved; then a control gain is dynamically mapped according to the modulus, and an integral link is constrained by using a body temperature change rate data, so as to output a safety regulation instruction; a system built-in parameter baseline back-off logic is provided; through dynamic mapping of a running state to a control parameter space, the application realizes mode conversion of power equipment from passive over-limit removal to active energy envelope regulation, effectively solves the mutual exclusion contradiction between regulation response bandwidth and steady-state accuracy, and improves the nonlinear disturbance receiving capacity of a hub node under a 50Hz fundamental frequency.

Description

Technical Field

[0001] This invention relates to a power equipment safety regulation system based on data analysis, belonging to the field of power system automation control technology. Background Technology

[0002] Current power equipment safety regulation systems are used to ensure that the physical state of the equipment is within a stable envelope. The mainstream solution adopts a closed-loop feedback regulation method, which compares the current and voltage data collected by sensors with the set reference values, and calculates the regulation command according to the preset proportional, integral and derivative control algorithms to drive the actuator to generate compensation for the deviation. This type of solution has a simple structure and high reliability, and has been widely used in the automation control of conventional power systems. With the high proportion of new energy connected to the distribution network hub nodes, the load disturbances experienced by power equipment exhibit obvious nonlinear and time-varying characteristics. In control engineering practice, in order to maintain the regulation robustness under all operating conditions, existing technologies improve the equipment's anti-disturbance performance by improving the physical characteristics of components and optimizing the body shape. Hardware fine-tuning is limited by physical space and material strength, and it is difficult to cover complex operating conditions. Software control logic also faces challenges. For example, Chinese invention patent application with publication number CN120128510A discloses an intelligent network connection management and data transmission method and related equipment. The proportional gain and integral constant inside the controller are usually set as conservative static values ​​based on standard load conditions to avoid high-frequency oscillations of the control mechanism under small disturbances.

[0003] However, this static regulation method has performance incompatibilities when dealing with sudden high-frequency step disturbances. If a conservative low gain setting is maintained, the regulation loop will produce a significant phase response delay, causing the physical state of the equipment to exceed the limit for a short time before the action is completed, which may lead to inrush current or thermal runaway risks. If the feedback gain is increased globally to shorten the response settling time, the system is prone to regulation oscillations during the process of approaching steady state, which may cause mechanical fatigue or electrical losses in the actuators. Conventional improvement approaches in the industry often focus on increasing hardware redundancy or paralleling multiple independent regulators. This not only increases the complexity of the system components, but also fails to solve the structural contradiction between dynamic disturbance rejection and steady-state robustness from the perspective of regulation law reconstruction.

[0004] Therefore, the technical problem to be solved by this invention is how to establish a regulation mechanism that can sense the evolution of the manifold in the operating state and reconstruct the internal structure of the control law in real time, so as to achieve high bandwidth suppression of transient disturbances while ensuring the robustness of the closed loop of the system. Summary of the Invention

[0005] To address the problems in the background art, the technical solution of the present invention is as follows: A power equipment safety regulation system based on data analysis, comprising: The data acquisition unit is used to acquire the real-time output status sequence of the power equipment and the temperature change rate data of the equipment itself. The state space parsing unit is used to construct the operating state space based on the real-time output state sequence and calculate the magnitude of the state offset vector of the current state point of the power equipment relative to the preset safety boundary in the operating state space. The nonlinear adaptive adjustment unit is used to generate safety adjustment commands based on the magnitude of the state offset vector and the body temperature change rate data. Its internal logic procedure includes the following steps: Step S11, establishing a correlation between the control gain and the magnitude of the state offset vector according to a preset gain mapping function, so that the control gain increases as the magnitude of the state offset vector decreases; Step S12, superimposing the body temperature change rate data as an integral feedback constraint factor into the control loop to suppress the accumulation rate of the integral term; Step S13, performing closed-loop calculation based on the target control gain calculated according to the gain mapping function and the control loop with the integral feedback constraint factor, and outputting the safety adjustment command to the execution component of the power equipment; The nonlinear adaptive adjustment unit has built-in parameter baseline fallback logic, which is used to monitor the time interval of receiving the target control gain, and when the time interval exceeds a preset communication heartbeat threshold, drive the currently effective target control gain to return to the static reference parameter according to a preset exponential decay curve.

[0006] Preferably, the state space parsing unit includes the following sub-steps when calculating the magnitude of the state offset vector: Step S21, reconstruct the phase space of the real-time output state sequence to obtain a multi-dimensional state vector characterizing the transient fluctuations of the power equipment; Step S22, fit the operating trajectory in the operating state space using the multi-dimensional state vector and determine the boundary surface of the safe operating threshold in the operating state space; Step S23, calculate the minimum geometric distance between the multi-dimensional state vector and the boundary surface at the sampling time, and define the minimum geometric distance as the magnitude of the state offset vector.

[0007] Preferably, the nonlinear adaptive adjustment unit in step S11 is associated with the following sub-mechanisms: step S111, real-time monitoring of the dynamic change rate of the state offset vector magnitude, and activation of the proportional compensation logic when the running state point approaches the boundary surface; step S112, adjusting the weight coefficient of the proportional adjustment link according to the inverse step characteristic of the state offset vector magnitude to enhance the system's ability to suppress high-frequency disturbances.

[0008] Preferably, in step S12, the constraint path of the integral feedback constraint factor in the nonlinear adaptive adjustment unit is as follows: Step S121, convert the body temperature change rate data into constraint weights characterizing the temperature rise risk of the power equipment; Step S122, introduce the constraint weights as a penalty term into the integral operator of the control loop, and when the body temperature change rate data exceeds the preset thermal response threshold, reduce the integral action through the constraint weights to prevent thermal runaway caused by adjustment overshoot.

[0009] Preferably, the system includes a link quality monitoring unit, which is used to acquire data transmission quality parameters between the data acquisition unit and the nonlinear adaptive adjustment unit; when the data transmission quality parameters are lower than a preset communication reliability threshold, the nonlinear adaptive adjustment unit initiates parameter baseline fallback logic to return the target control gain to the static reference parameter according to the exponential decay curve.

[0010] Preferably, the data acquisition unit includes a power acquisition component and an environmental sensing component; the power acquisition component is used to... The sampling frequency is synchronized to acquire current phasor data and instantaneous voltage value data of the power equipment and encapsulate them into a real-time output state sequence; the environmental sensing component is used to acquire temperature data through a sensor array deployed at the heat dissipation nodes of the power equipment and perform first-order derivative processing on the temperature data to generate body temperature change rate data in units of K / s.

[0011] Preferably, the system stores an adjustment mode database, which records preset safe operation characteristic curves for different operating conditions; the nonlinear adaptive adjustment unit identifies the current operating condition type based on the geometric topological features of the operating state space and retrieves the corresponding safe operation threshold from the adjustment mode database.

[0012] Preferably, the nonlinear adaptive adjustment unit performs an instruction consistency check before outputting the safety adjustment instruction: comparing the expected response amount corresponding to the current safety adjustment instruction with the physical output range of the execution component; when the expected response amount exceeds the physical output range, the safety adjustment instruction is saturated and limited according to the physical output range.

[0013] Preferably, a parameter coupling link is established between the state space analysis unit and the nonlinear adaptive adjustment unit; the system uses the parameter coupling link to map the real-time data space to the parameter space of the control loop, in order to improve the nonlinear disturbance acceptance capability of the power equipment at the 50Hz fundamental frequency.

[0014] Compared with the prior art, the beneficial effects of the present invention are: Firstly, in the safe regulation of power equipment, a decoupled two-layer regulation topology is adopted. By utilizing the coordinated interaction between the fast control inner loop and the slow parameter outer loop, the mutual exclusion relationship between regulation response lag and steady-state oscillation in traditional fixed-gain control is resolved. This enables the power equipment to generate a control law that matches the current disturbance frequency under different load conditions, compresses the establishment time of regulation commands, eliminates the short-term physical state overshoot caused by conservative gain, and avoids mechanical or electrical wear caused by over-regulation when the system approaches steady state.

[0015] Secondly, by calculating the normal distance from the operating state trajectory to the safety envelope boundary in the state phase space, the control gain is nonlinearly drifted with the degree of deviation of the equipment's operating state. This state manifold-based driving method changes the limitations of single error driving. When the equipment state deviates from the warning threshold, the nonlinear proportional compensation weight is exponentially enhanced as the normal distance decreases. Thus, without changing the physical characteristics of the actuator, the dynamic response bandwidth of the control system is broadened, and the system's ability to suppress nonlinear disturbances is enhanced.

[0016] Third, the temperature change rate of the equipment body is introduced as a physical damping constraint of the integral term. Combined with the parameter smoothing and fallback mechanism under communication status monitoring, an endogenous defense system against control loop divergence is constructed. By superimposing a damping constraint factor in the integral term, the system can instantly suppress over-adjustment under the condition of a surge in heat load, block the trend of thermal runaway, and, together with the smooth degradation logic of the parameter baseline when communication is abnormal, ensure that the control system can smoothly return from dynamic gain to static reference parameters under complex electromagnetic interference or data loss real environment, and maintain the absolute stability of the equipment's safe operating envelope margin. Attached Figure Description

[0017] Figure 1 This invention relates to the system dynamic control logic and data flow diagram; Figure 2 This invention relates to a block diagram of the system functional units and their interaction relationships. Detailed Implementation

[0018] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0019] A data analysis-based power equipment safety regulation system includes: The data acquisition unit is used to acquire the real-time output status sequence of the power equipment and the temperature change rate data of the equipment itself. The state space parsing unit is used to construct the operating state space based on the real-time output state sequence and calculate the magnitude of the state offset vector of the current state point of the power equipment relative to the preset safety boundary in the operating state space. The nonlinear adaptive adjustment unit is used to generate safety adjustment commands based on the magnitude of the state offset vector and the body temperature change rate data. Its internal logic procedure includes the following steps: Step S11, establishing a correlation between the control gain and the magnitude of the state offset vector according to a preset gain mapping function, so that the control gain increases as the magnitude of the state offset vector decreases; Step S12, superimposing the body temperature change rate data as an integral feedback constraint factor into the control loop to suppress the accumulation rate of the integral term; Step S13, performing closed-loop calculation based on the target control gain calculated according to the gain mapping function and the control loop with the integral feedback constraint factor, and outputting the safety adjustment command to the execution component of the power equipment; The nonlinear adaptive adjustment unit has built-in parameter baseline fallback logic, which is used to monitor the time interval of receiving the target control gain, and when the time interval exceeds a preset communication heartbeat threshold, drive the currently effective target control gain to return to the static reference parameter according to a preset exponential decay curve.

[0020] Preferably, the state space parsing unit includes the following sub-steps when calculating the magnitude of the state offset vector: Step S21, reconstruct the phase space of the real-time output state sequence to obtain a multi-dimensional state vector characterizing the transient fluctuations of the power equipment; Step S22, fit the operating trajectory in the operating state space using the multi-dimensional state vector and determine the boundary surface of the safe operating threshold in the operating state space; Step S23, calculate the minimum geometric distance between the multi-dimensional state vector and the boundary surface at the sampling time, and define the minimum geometric distance as the magnitude of the state offset vector.

[0021] Preferably, the nonlinear adaptive adjustment unit in step S11 is associated with the following sub-mechanisms: step S111, real-time monitoring of the dynamic change rate of the state offset vector magnitude, and activation of the proportional compensation logic when the running state point approaches the boundary surface; step S112, adjusting the weight coefficient of the proportional adjustment link according to the inverse step characteristic of the state offset vector magnitude to enhance the system's ability to suppress high-frequency disturbances.

[0022] Preferably, in step S12, the constraint path of the integral feedback constraint factor in the nonlinear adaptive adjustment unit is as follows: Step S121, convert the body temperature change rate data into constraint weights characterizing the temperature rise risk of the power equipment; Step S122, introduce the constraint weights as a penalty term into the integral operator of the control loop, and when the body temperature change rate data exceeds the preset thermal response threshold, reduce the integral action through the constraint weights to prevent thermal runaway caused by adjustment overshoot.

[0023] Preferably, the system includes a link quality monitoring unit, which is used to acquire data transmission quality parameters between the data acquisition unit and the nonlinear adaptive adjustment unit; when the data transmission quality parameters are lower than a preset communication reliability threshold, the nonlinear adaptive adjustment unit initiates parameter baseline fallback logic to return the target control gain to the static reference parameter according to the exponential decay curve.

[0024] Preferably, the data acquisition unit includes a power acquisition component and an environmental sensing component; the power acquisition component is used to... The sampling frequency is synchronized to acquire current phasor data and instantaneous voltage value data of the power equipment and encapsulate them into a real-time output state sequence; the environmental sensing component is used to acquire temperature data through a sensor array deployed at the heat dissipation nodes of the power equipment and perform first-order derivative processing on the temperature data to generate body temperature change rate data in units of K / s.

[0025] Preferably, the system stores an adjustment mode database, which records preset safe operation characteristic curves for different operating conditions; the nonlinear adaptive adjustment unit identifies the current operating condition type based on the geometric topological features of the operating state space and retrieves the corresponding safe operation threshold from the adjustment mode database.

[0026] Preferably, the nonlinear adaptive adjustment unit performs an instruction consistency check before outputting the safety adjustment instruction: comparing the expected response amount corresponding to the current safety adjustment instruction with the physical output range of the execution component; when the expected response amount exceeds the physical output range, the safety adjustment instruction is saturated and limited according to the physical output range.

[0027] Preferably, a parameter coupling link is established between the state space analysis unit and the nonlinear adaptive adjustment unit; the system uses the parameter coupling link to map the real-time data space to the parameter space of the control loop, in order to improve the nonlinear disturbance acceptance capability of the power equipment at the 50Hz fundamental frequency.

[0028] Example 1: In the operation scenario of a power distribution network hub node with a renewable energy penetration rate exceeding 40%, power equipment faces high-frequency and nonlinear source-load dual-side disturbances. Due to the microsecond-level jumps in distributed photovoltaic output with ambient irradiance, conventional regulation loops following static error feedback exhibit response delays. If the factory-set conservative gain is maintained, the system experiences phase lag when responding to transient energy surges, causing the equipment's physical state to briefly exceed the preset safe operating boundary, triggering the risk of inrush current. Under this operating condition, the data acquisition unit acquires current phasor data and instantaneous voltage values ​​of the power equipment at a 1μs sampling frequency and encapsulates them into a real-time output state sequence. The state space analysis unit receives this sequence and performs phase space reconstruction to obtain the characterization... The multidimensional state vector of transient fluctuations in power equipment is fitted with the operating trajectory within the operating state space, and the boundary surface for determining the safe operating threshold is determined. The minimum geometric distance between the multidimensional state vector and the boundary surface at the sampling time is calculated and defined as the magnitude of the state offset vector, where 1 μs is the sampling period, used to define the time resolution of data acquisition. The nonlinear adaptive adjustment unit establishes a correlation between the control gain and the magnitude of the state offset vector according to a preset gain mapping function, so that the control gain increases as the magnitude of the state offset vector decreases. This dynamic scheduling mechanism based on the state phase space deviation enables the fast control inner loop to automatically activate high-weight proportional compensation the instant the system approaches the safe boundary, compressing the adjustment response settling time to the limit of physical component action. To address the mutually exclusive relationship between dynamic response bandwidth and steady-state reliability in conventional static control, this invention addresses the issue of frequent actuator movements driven by control commands during high-frequency step disturbances. These movements generate alternating electrical switching losses and Joule heat, which are the physical causes of core temperature rise. The inherent second-level time delay in the physical heat conduction process at heat dissipation nodes means that directly using surface temperature data to constrain the microsecond-level electrical inner loop inevitably leads to control instability. To resolve this cross-scale mismatch, this invention constructs a transmission chain from energy flow to physical entities: transient Joule heat power is calculated in real-time using instantaneous current values ​​obtained through microsecond-level sampling, serving as the microscopic source term for heat generation; a first-order thermal resistance-capacitance network model simulates the heat conduction path from the internal core to external heat dissipation nodes. This process transforms rapidly changing electrical... The gas stress is transformed into a thermal accumulation trend prediction, allowing the slow temperature change rate to be used as a soft predictive boundary to intervene in the control logic in advance. Specifically, when the predicted thermal power accumulation rate indicates that the core will approach the thermal failure point within milliseconds, the system reduces the weight of the integral operator to suppress the output intensity of the adjustment command. This introduces physical thermodynamic damping into the microsecond-level electrical control loop, eliminating the spatiotemporal mismatch of multi-physics fields across scales. The system establishes a thermal delay compensation procedure based on the first-order thermal resistance-capacitance network physical model of heat transfer, extracts the instantaneous current value from the real-time output state sequence to calculate the transient heating power, and constructs a state observer by combining the historical temperature rise data of sensor nodes to compensate for the phase lag of the measured data and predict the true temperature evolution of the core inside the device.

[0029] When nonlinear disturbances cause an increase in internal heat loss of the equipment, the environmental sensing component acquires temperature data through a sensor array deployed at the heat dissipation nodes of the power equipment and performs a first-order derivative operation to generate a unit temperature value. The nonlinear adaptive control unit converts the body temperature change rate data into constraint weights characterizing the temperature rise risk of the power equipment, based on the exponentially decaying damping physical model. Constraint weights According to the analytical expression Calculated and obtained, among which, parameters Refers to the rate of change of body temperature, a constant. To limit the thermal response start-up threshold to the range of 0.5 K / s to 1.2 K / s, the coefficient The operator is dimensionlessly calibrated as the thermodynamic decay constant s / K. In engineering calibration, the maximum value calculation is indicated. The value is taken as 0.5 to 1.5 times the reciprocal of the rated temperature rise rate of the equipment. Its physical meaning lies in quantifying the sensitivity of integral regulation to temperature rise risk: when When the value is set too small, the temperature-constrained decay of the integral term is insufficient, making it difficult to prevent thermal runaway; when... When the overshoot is too large, normal adjustment actions will be excessively suppressed by small temperature fluctuations, leading to a decrease in steady-state accuracy. In actual operation, by monitoring the transition time from thermal response start-up to steady state of the equipment under constant load abrupt change conditions, the critical value that minimizes the accumulation of overshoot heat is selected as the final value. Calibration parameters, control loop based on mathematical model Output closed-loop integral control quantity, parameter Identify the integral adjustment output component, constant Represents the hardware setting of the basic integral gain, variable The instantaneous state offset is calculated for the current moment and introduced into the integral operator of the control loop as a penalty term. When the temperature change rate of the device exceeds the preset thermal response threshold, the integral action is reduced by the constraint weight to block the thermal runaway trend caused by the regulation overshoot. At the same time, the link quality monitoring unit obtains the data transmission quality parameters between the data acquisition unit and the nonlinear adaptive adjustment unit. When the parameter is lower than the preset communication reliability threshold or the time interval for receiving the target control gain exceeds the preset communication heartbeat threshold, the parameter baseline fallback logic is triggered, driving the currently effective target control gain to return to the static reference parameter according to the preset exponential decay curve. After the power equipment undergoes continuous high-frequency step disturbance test, its output state is maintained within the preset energy envelope. The establishment time of the adjustment command is reduced by more than 60% compared with the fixed parameter mode. The maximum temperature of the device is controlled below the safety threshold. The system automatically degrades to the static reference parameter operation under the simulated communication interruption condition, realizing the transformation of the power equipment from passive over-limit cutoff to active energy envelope regulation.

[0030] Example 2: Verification of the safety regulation performance of power equipment was conducted in a hybrid analog-digital physics experimental platform. An electromagnetic transient model simulating a 10kV distribution transformer and a programmable electronic load array were used to construct the operating environment. The data acquisition unit obtained the real-time output state sequence through a 16-bit analog-to-digital conversion path. The experimental data was collected from the original operating waveforms of the simulated distribution network hub node under high-proportion distributed photovoltaic access conditions. To simulate the industrial electromagnetic environment, Gaussian white noise with a signal-to-noise ratio of 20dB and power frequency interference harmonics at a frequency of 50Hz were superimposed on the current and voltage signals at the acquisition end. The sampling period setting needed to balance the time resolution of data acquisition with the computational load of state space reconstruction. If the sampling period was greater than 1μs, signal aliasing would occur during state space trajectory fitting, leading to a decrease in the recognition accuracy of the safety boundary surface. If the sampling period was less than 1μs, the computational delay of data parsing would accumulate in the regulation loop and induce a decrease in closed-loop stability. Therefore, the sampling period was determined to be 1μs to capture microsecond-level transient excitation characteristics at the 50Hz fundamental frequency.

[0031] When inputting raw current phasor and voltage instantaneous values ​​containing 20dB Gaussian white noise, the state space analysis unit reconstructs the phase space of the real-time output state sequence to obtain a multidimensional state vector. The experimental group of this invention calculates the magnitude of the state offset vector in real time through the geometric mapping of the running trajectory and the boundary surface. The control group removes this analysis step and adopts a gain control method based on fixed deviation feedback. When the simulated electronic load generates a 50% load step change, the original error signal detected by the control group fluctuates due to noise interference, causing the adjustment command to generate 4 polarity switches within 12.8ms. Its adjustment response settling time is 15.2ms and accompanied by a decaying oscillation lasting 4.5ms. The experimental group extracts a smooth state offset vector magnitude sequence in the noisy environment through phase space dimensionality reduction processing. It is observed that the magnitude decreases from 4.52 in the initial steady state to 0.78 near the boundary. This quantification result of the physical state deviation triggers the nonlinear adaptive adjustment unit to increase the control gain, compressing the adjustment response settling time to 4.1ms, and the overshoot of the output state waveform is reduced by 72% compared with the control group.

[0032] To verify the regulation law of the scheme under thermodynamic constraints, a gradient test of the body temperature change rate was set in the prototype of this invention. When the body temperature change rate output by the environmental sensing component increased from 0.2 K / s to 2.5 K / s, the nonlinear adaptive adjustment unit converted the change rate into an integral feedback constraint factor. It was observed that the weight of the integral operator in the control loop decreased with the increase of the temperature rise rate. This dynamic constraint mechanism suppressed the overshoot heat accumulation caused by integral saturation, keeping the maximum temperature of the equipment within a stable range of 65.2℃. As an out-of-range control group, when the body temperature change rate was artificially set to an overload condition of 5.5 K / s, a performance inflection point was observed. At this time, the constraint weight reached the saturation upper limit of 1.0, and the system cut off the integral link and returned to the reference power. The operating mode prevents equipment damage. In the simulated communication interference test, the link quality monitoring unit detected that the time interval of the received target control gain exceeded the communication heartbeat threshold of 100ms. At this time, the parameter baseline fallback logic was triggered, and the effective control gain fell back from the dynamic gain to the static reference parameter according to the preset exponential decay curve, maintaining the controlled state of the system. The adjustment command establishment time of the sample group under nonlinear load impact was reduced by more than 60% compared with the control group. Moreover, it exhibited nonlinear compensation characteristics and physical safety boundary constraints under 20dB noise interference and gradient temperature rise conditions. The output state of the power equipment was maintained within the preset energy envelope, realizing the transformation of the power equipment from traditional static error compensation to dynamic envelope adjustment based on phase space trajectory.

[0033] Example 3: This example implements short-circuit fault current limiting regulation at a DC distribution network hub node including flexible interconnected switches. Due to network-induced delays in control command transmission, if the control gain mapping relationship of the nonlinear adaptive regulation unit is not aligned with the physical delay, the regulation amount will induce oscillations at the moment the system state dynamically exceeds the limit, causing the voltage stress on the power equipment to exceed the withstand limit of the semiconductor device. Under this condition, the state space parsing unit obtains the following information: Voltage and current sequences at each sampling point, where Given the sample length; to eliminate dimensional redundancy in phase space reconstruction, the system follows a calibration procedure based on the autocorrelation function, calculating the sequence autocorrelation coefficient to decrease to its initial value. The time interval is determined as the delay time. Determining the embedding dimension using the spurious nearest neighbor method The system monitors the rate of change of the nearest neighbor distance after increasing the dimension. When the rate of change is less than 5%, the dimension increase is stopped, thus constructing a three-dimensional operating state space. The state space parsing unit uses the set of historical normal operating trajectory points in the three-dimensional operating state space to complete incremental convex hull fitting to generate a safe boundary surface. Under the constraint of a 1μs sampling period, the incremental convex hull fitting adopts a recursive local update strategy. The system maintains a vertex set composed of the current normal operating sample points. When the newly generated state trajectory point is inside the original convex hull, no structural update is performed. If the new trajectory point is offset, the system only reconstructs the boundary surface in the neighborhood of that point. By calculating the spatial orientation relationship between the new point and the adjacent triangular facets, the old vertices that are covered are removed and new facet connections are established. Thus, the dynamic evolution of the boundary surface is achieved without traversing the entire dataset. This local update logic controls the computational complexity of a single parsing to the constant level, ensuring the synchronization of the boundary parsing and the microsecond-level sampling process. The minimum geometric distance between the multidimensional state vector and the safe boundary surface at the sampling time is calculated, and this minimum geometric distance is determined as the magnitude of the state offset vector.

[0034] The nonlinear adaptive control unit is based on the maximum allowable rise time of the power equipment. The unit establishes the control gain by determining the gain mapping function. The control gain is inversely proportional to the magnitude of the state offset vector. As the magnitude of the state offset vector decreases, the adjustment follows a quadratic function trend. When the rate of change of the body temperature exceeds 0.5 K / s, the nonlinear adaptive adjustment unit converts this rate of change data into constraint weights characterizing the temperature rise risk of the power equipment and adds them to the control loop to suppress the accumulation rate of the integral term. The system determines the communication heartbeat threshold based on the open-loop gain margin of the closed-loop system. It is set to 5 to 8 times the sampling period of the adjustment loop, when the time interval for receiving the target control gain exceeds At that time, the parameter baseline fallback logic drives the currently effective target control gain according to the formula. Return, among which, for Instantaneous gain at time, The original gain at the moment of failure. The damping coefficient is determined based on the transient damping ratio during the system degradation process. For communication heartbeat threshold, The attenuation coefficient is... The time is calculated from the moment of communication interruption; under the drive of the parameter-calibrated regulation system, the phase margin of the regulation response of the power equipment is not less than 45 degrees. In a communication environment simulating a 10% packet loss rate, the system maintains a controlled operation capability of 500ms through parameter baseline fallback logic. The turn-off overvoltage of semiconductor devices is controlled within 1.1 times the rated voltage, realizing stable control of the power equipment safety regulation system under time delay constraints.

[0035] Example 4: In the commissioning scenario of adaptive parameter configuration for DC circuit breakers, the nonlinear adaptive adjustment unit calibrates the distribution coefficient of the gain mapping function through a standard pulse excitation response experiment. This procedure includes applying a step interference signal with an amplitude of 10% of the rated value to the power equipment under controlled conditions and recording different control gains. System overshoot and adjusting response settling time The recorded response dataset is fitted to a preset gain mapping function using the least squares method. The weight distribution coefficients of the gain mapping function are determined based on the physical constraint that the response bandwidth should not be less than 500Hz, thus controlling the gain. The functional relationship curve between the state offset vector magnitude and the actuator's dead zone and saturated nonlinearity is physically aligned with these characteristics.

[0036] When the system faces a sampling zero-point drift caused by sensor aging, the data acquisition unit completes the baseline calibration procedure within a preset window when the power equipment enters maintenance mode. This procedure reads the residual level values ​​of each sampling channel in the zero-input state and calculates them as a static bias vector. In the real-time data processing stream, the static bias vector is subtracted from the current original sampling vector. To obtain the corrected real-time output state sequence, and at the same time use the zero-crossing synchronization signal of the fundamental frequency of the power system as a time reference, phase alignment compensation is performed on the waveforms collected by each sensor to ensure that the components of the multidimensional state vector have consistency at the time domain sampling points when the state space analysis unit constructs the operating state space, suppressing the turn-off overvoltage of the semiconductor device to within 1.1 times the rated voltage, and maintaining the stable state of the power equipment safety regulation system under signal drift constraints.

[0037] Example 5: In a distributed cluster coordinated regulation scenario involving multiple intelligent transformers operating in parallel, the system receives a large power regulation command from the dispatch center. The nonlinear adaptive regulation unit advances the cluster coordinated management procedure, determining the distribution weight of the regulation task by obtaining the magnitude of the state offset vector of each power device in the cluster. This procedure establishes the regulation weight. It is positively correlated with the magnitude of the state offset vector, where, For the first Adjustment weights of the equipment; adjustment weights The calculation follows the normalization criterion, which allocates higher power regulation increments to power equipment with larger state offset vector magnitudes, while power equipment with smaller state offset vector magnitudes maintains the baseline power operation state to avoid the safe operation threshold. This resource allocation logic based on the physical safety manifold enables the adjustment stress of each node to adaptively adjust with its safety margin when the cluster responds to power step demand, and the cluster frequency deviation is controlled within 0.02Hz.

[0038] When the system faces sensor measurement zero-point drift induced by ambient temperature cycling, the nonlinear adaptive adjustment unit advances the reference calibration procedure within the preset maintenance window of the power equipment, and the data acquisition unit acquires the residual level value under zero input state and determines it as the static bias vector. The system calculates the static bias vector. The zero-point compensation parameters of the local memory are updated by taking the average value over 10 consecutive sampling periods, and the updated parameters are embedded as a subtraction operator into the real-time data processing stream to eliminate cumulative measurement deviations. This timeliness guarantee mechanism for dynamic components ensures the geometric coordinate fidelity of the operating state space during long-term service. After 5000 hours of continuous operation, the analytical error of the state offset vector magnitude of the power equipment is controlled within 1%, maintaining the stability of the power equipment safety regulation system throughout its entire life cycle.

[0039] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0040] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A power equipment safety regulation system based on data analysis, characterized in that, include: The data acquisition unit is used to acquire the real-time output status sequence of the power equipment and the temperature change rate data of the equipment itself. The state space parsing unit is used to construct the operating state space based on the real-time output state sequence and calculate the magnitude of the state offset vector of the current state point of the power equipment relative to the preset safety boundary in the operating state space. The nonlinear adaptive control unit is used to generate a safety control command based on the magnitude of the state offset vector and the body temperature change rate data. Its internal logic procedure includes the following steps: Step S11, establishing a relationship between the control gain and the magnitude of the state offset vector according to a preset gain mapping function, so that the control gain increases as the magnitude of the state offset vector decreases; Step S12, superimposing the body temperature change rate data as an integral feedback constraint factor into the control loop to suppress the accumulation rate of the integral term; Step S13, performing closed-loop calculation based on the target control gain calculated by the gain mapping function and the control loop with the integral feedback constraint factor, and outputting the safety control command to the execution component of the power equipment. The nonlinear adaptive adjustment unit has built-in parameter baseline fallback logic, which is used to monitor the time interval of the received target control gain, and when the time interval exceeds the preset communication heartbeat threshold, it drives the currently effective target control gain to return to the static reference parameter according to the preset exponential decay curve.

2. The power equipment safety regulation system based on data analysis according to claim 1, characterized in that, When the state space parsing unit calculates the magnitude of the state offset vector, it includes the following sub-steps: Step S21, reconstruct the phase space of the real-time output state sequence to obtain a multi-dimensional state vector characterizing the transient fluctuations of the power equipment; Step S22: Fit the running trajectory in the running state space using the multidimensional state vector, and determine the boundary surface of the safe running threshold in the running state space; Step S23: Calculate the minimum geometric distance of the multidimensional state vector pointing to the boundary surface at the sampling time, and define the minimum geometric distance as the magnitude of the state offset vector.

3. The power equipment safety regulation system based on data analysis according to claim 1, characterized in that, In step S11, the nonlinear adaptive adjustment unit is associated with the following sub-mechanisms: Step S111, real-time monitoring of the dynamic rate of change of the magnitude of the state offset vector, and activation of the proportional compensation logic when the running state point approaches the boundary surface; Step S112, based on the inverse step characteristic of the magnitude of the state offset vector, increasing the weight coefficient of the proportional adjustment link to enhance the system's ability to suppress high-frequency disturbances.

4. The power equipment safety regulation system based on data analysis according to claim 1, characterized in that, In step S12, the constraint path of the integral feedback constraint factor in the nonlinear adaptive adjustment unit is as follows: Step S121, convert the body temperature change rate data into constraint weights that characterize the temperature rise risk of the power equipment; Step S122, introduce the constraint weights as a penalty term into the integral operator of the control loop, and when the body temperature change rate data exceeds the preset thermal response threshold, reduce the integral action through the constraint weights to prevent thermal runaway caused by adjustment overshoot.

5. The power equipment safety regulation system based on data analysis according to claim 1, characterized in that, The system includes a link quality monitoring unit, which is used to acquire data transmission quality parameters between the data acquisition unit and the nonlinear adaptive adjustment unit. When the data transmission quality parameters are lower than the preset communication reliability threshold, the nonlinear adaptive adjustment unit initiates the parameter baseline fallback logic to return the target control gain to the static reference parameter according to the exponential decay curve.

6. The power equipment safety regulation system based on data analysis according to claim 1, characterized in that, The data acquisition unit includes a power acquisition component and an environmental sensing component; the power acquisition component is used to... The sampling frequency is synchronized to acquire current phasor data and instantaneous voltage value data of the power equipment and encapsulate them into a real-time output state sequence; the environmental sensing component is used to acquire temperature data through a sensor array deployed at the heat dissipation nodes of the power equipment and perform first-order derivative processing on the temperature data to generate body temperature change rate data in units of K / s.

7. The power equipment safety regulation system based on data analysis according to claim 1, characterized in that, The system stores an adjustment method database, which records preset safe operation characteristic curves for different operating conditions. The nonlinear adaptive control unit identifies the current operating condition type based on the geometric topological features of the operating state space and retrieves the corresponding safe operating threshold from the control method database.

8. A power equipment safety regulation system based on data analysis according to claim 1, characterized in that, Before outputting a safety adjustment command, the nonlinear adaptive adjustment unit performs a command consistency check: it compares the expected response amount corresponding to the current safety adjustment command with the physical output range of the execution component; when the expected response amount exceeds the physical output range, it saturates and limits the safety adjustment command according to the physical output range.

9. A power equipment safety regulation system based on data analysis according to claim 1, characterized in that, A parameter coupling link is established between the state space analysis unit and the nonlinear adaptive adjustment unit; the system uses the parameter coupling link to map the real-time data space to the parameter space of the control loop, which is used to improve the nonlinear disturbance acceptance capability of power equipment at the 50Hz fundamental frequency.

Images