An intelligent diagnosis and energy efficiency optimization system for power facility material degradation detection

Abstract

The application provides a kind of intelligent diagnosis and energy efficiency optimization system for power facility material degradation detection, it is related to detection management field, including: by state summary extraction module obtains and pre-processes field state summary;Utilize the execution module of gate frozen to execute multiple rounds of confirmation, drift freezing and attribution delay processing, determine material degradation attribution state by attribution reset;Dispatch operation control module matches dispatch rule and capacity balance constraint according to attribution result, and the power facility is divided into corresponding operation pool to execute dispatch task;Feedback parameter correction module captures control defects and generates correction label by analyzing state summary and load bearing result, and adjusts threshold and control boundary with variable step size by combining decision damping and cross confirmation operation;The application can prevent material degradation misjudgment and inhibit the determination parameter shock, realizes the accurate diagnosis of power facility and the reliable scheduling of energy efficiency.

Description

Technical Field

[0001] This invention relates to the field of testing and management, specifically to an intelligent diagnostic and energy efficiency optimization system for detecting material degradation in power facilities. Background Technology

[0002] With the expansion of power system scale and the increasing complexity of operating environment, the health of power facility materials directly affects the stability of energy transmission. Currently, the power industry is accelerating its transformation from manual inspection to automated diagnosis. In the context of digitalization, real-time monitoring and assessment of equipment material deterioration has become a key link in ensuring power grid safety. This trend is driving power operation and maintenance to evolve towards preventive management and in-depth evolution throughout the entire life cycle.

[0003] Existing technologies typically employ online monitoring schemes based on fixed thresholds to manage power facilities. This involves collecting real-time physical quantities through sensors and comparing them with preset standard ranges. When the collected data exceeds a specific limit, the system immediately triggers an early warning signal and automatically switches to the corresponding maintenance procedure. This approach mainly relies on empirical model settings and uses simple logical discrimination to classify and process equipment conditions.

[0004] The shortcomings of existing technologies are: when responding to external interference signals, existing methods are difficult to balance the immediacy and accuracy of identification; when faced with uncertain signs, the lack of intermediate layer inspection and retention mechanisms can easily lead to deviations in the final evaluation; and fixed control logic cannot cope with causal confusion in complex scenarios, which restricts the reliability of atypical transformation processes. Summary of the Invention

[0005] (a) Technical problems to be solved

[0006] To address the shortcomings of existing technologies, this invention provides an intelligent diagnostic and energy efficiency optimization system for detecting material degradation in power facilities. This system addresses the issues raised in the background section, such as the difficulty of balancing immediacy and accuracy in responding to external interference signals, the lack of intermediate-level inspection and retention mechanisms in dealing with uncertain signs leading to biased final evaluations, and the inability of fixed control logic to handle causal confusion in complex scenarios.

[0007] (II) Technical Solution

[0008] To achieve the above objectives, the present invention is implemented through the following technical solution: an intelligent diagnostic and energy efficiency optimization system for detecting material degradation in power facilities, comprising a status summary extraction module, a gating freeze execution module, a scheduling operation control module, and a feedback parameter correction module;

[0009] The status summary extraction module obtains the on-site status summary of the power facilities, and transmits the on-site status summary to the gated freeze execution module after performing feature normalization processing.

[0010] The gated freeze execution module receives the field status summary and determines whether the field status summary meets the multi-round confirmation conditions based on the operating conditions. If the field status summary does not meet the multi-round confirmation conditions, the field status summary is classified into an unconfirmable deterioration state and pushed into the drift freeze layer to retain the observation data stream. If the field status summary meets the multi-round confirmation conditions, a delay attribution window is allocated to the field status summary and it enters the second-level attribution delay processing structure. At the same time, the continuous identical attribution operation is prohibited to form a single stable judgment, thereby determining the corresponding attribution state and triggering the corresponding material deterioration maintenance chain. The attribution state is then transmitted to the scheduling operation control module.

[0011] The scheduling and operation control module matches the corresponding scheduling permission rules and capacity balance constraints according to the attribution state, divides the power facilities into stable state, undefined drift state, frozen state or pending review state, and then assigns them to the corresponding operation pool or conservative operation queue to allocate scheduling tasks, and transmits the load carrying results and state transition data to the feedback parameter correction module.

[0012] The feedback parameter correction module receives the load carrying results and the state transition data. It captures control failure samples by analyzing the self-fading of the field state summary, or determines delay control defects by analyzing the critical state of the power facilities after being subjected to the scheduling task. Based on this, it generates premature stability tendency labels or delay attribution insufficient labels, and adjusts the preset thresholds and judgment thresholds in the gating freeze execution module and the control boundaries in the scheduling operation control module by incremental adjustment of the step size through decision damping operation and cross-confirmation operation.

[0013] Preferably, the status summary extraction module, as the core of the perception layer of the system, acquires real-time field status summaries through current transformers, voltage sensors, and partial discharge detectors deployed at key locations of the power facilities. The underlying operations performed by the status summary extraction module first convert the analog signal into a digital sequence with a sampling frequency of 12000 Hz using an analog-to-digital converter. Subsequently, the status summary extraction module performs a Fast Fourier Transform on the digital sequence to extract characteristic frequency components, and encapsulates the extracted effective current value, harmonic distortion rate, and insulation dielectric loss tangent into a data packet representing the field status summary. During feature normalization, the status summary extraction module calls a preset min-max normalization algorithm to map all original physical parameters to a per-unit value range between 0 and 1. Specifically, this is achieved by calculating the difference between the current value and the minimum value of the measurement range, and dividing by the difference between the maximum and minimum values ​​of the measurement range, ensuring that power parameters of different dimensions are compared at the same order of magnitude. If the state summary extraction module detects that the input raw physical parameter exceeds 120% of the sensor's rated full scale, the state summary extraction module must perform a data truncation operation and trigger an abnormal data flag to prevent overflow errors from interfering with subsequent processing. After normalization, the state summary extraction module transmits the field state summary to the gated freeze execution module via an internal high-speed parallel bus, adding a globally unique timestamp index consisting of 17 bits to the transmission header. The state summary extraction module completes a full data scan cycle every 200 milliseconds. If a valid sensor feedback is not obtained within the scan cycle, the state summary extraction module forcibly fills the field state summary of the scan cycle with the value of the previous valid cycle and simultaneously sends a warning signal of data quality degradation to the gated freeze execution module. Through this combination of high-frequency sampling and strict normalization, the state summary extraction module provides a standardized input basis for subsequent material degradation diagnosis.

[0014] Preferably, after receiving the field status summary, the gated freeze execution module immediately calls its internal real-time comparator to execute the environmental phenomenon identification protocol. The gated freeze execution module compares the normalized amplitude feature term in the field status summary with the preset threshold of 0.15 point by point. If the field status summary is the first occurrence in the current observation sequence and its normalized amplitude is less than 0.15, the gated freeze execution module must determine the field status summary as the environmental phenomenon. In this case, the gated freeze execution module performs a mandatory path blocking operation, blocking the calculation path of the field status summary towards any material degradation conclusion, and strictly prohibiting the inclusion of such low-amplitude fluctuations in the damage accumulation model of the power facility. The gated freeze execution module redirects the environmental phenomenon to the drift freeze layer by modifying the routing control bit of the data packet. A temporary storage area is created and marked as the observation data stream. During the retention of the observation data stream, the gating freeze execution module restricts the system from triggering any form of equipment damage determination, and forcibly closes the maintenance action prompt interface for the power facility to ensure that the system does not generate false alarms due to background noise. If the field status summary is lower than the preset threshold of 0.15 for three consecutive sampling periods, the gating freeze execution module maintains the interception state. When the amplitude of the field status summary changes abruptly in a subsequent period and exceeds 0.15, the gating freeze execution module can release the blockade on the material degradation path, but the previous environmental phenomena still need to be retained as background parameters in the drift freeze layer for cross-comparison. Through this hard threshold-based interception mechanism, the gating freeze execution module cuts off the possibility of false alarms caused by environmental interference at the underlying logic level.

[0015] Preferably, the gated freeze execution module executes a refined unconfirmed degradation state management protocol for field status summaries that fail the initial screening. When the field status summary does not meet the multi-round confirmation conditions, the gated freeze execution module pushes the field status summary into the drift freeze layer using a first-in-first-out stack management mechanism to achieve long-term retention of unstable features. The underlying logic of the multi-round confirmation conditions stipulates that the field status summary must exhibit a monotonically increasing or stable feature offset within five consecutive sampling periods, and the offset increment in each period must be greater than 0.05, while its cumulative amplitude must be stable above 0.12. During the period when the field status summary is in the unconfirmed degradation state, the gated freeze execution module prohibits it from being called by the scheduling operation control module to ensure that unstable data does not participate in power distribution decisions. The drift freeze layer uses a circular buffer. The system continuously records subtle changes in the power facilities under disturbed conditions. If, after multiple rounds of filtering, the overall confidence index of the site status summary exceeds the judgment threshold of 0.85, the gated freeze execution module forms a single stability judgment. However, at the instant the single stability judgment is generated, the gated freeze execution module must execute a mandatory hierarchical blocking protocol, strictly prohibiting the single stability judgment from directly progressing to the first fault level. Instead, it forcibly diverts the data to the second-layer attribution delay processing structure for time-domain secondary verification. If the site status summary remains in the drift freeze layer for more than 48 hours and still fails to meet the multi-round confirmation conditions, the gated freeze execution module performs a data degradation operation, transferring the observed data stream to a non-volatile historical database and releasing the current active buffer space. This progressive confirmation mechanism ensures the prudence of the diagnostic conclusion.

[0016] Preferably, upon receiving the single stability judgment, the second-layer attribution delay processing structure immediately initiates the attribution verification process for the field status summary. The second-layer attribution delay processing structure first allocates a 48-hour delay attribution window to the field status summary, during which the attribution attribute of the field status summary is locked in a suspended state. During the open delay attribution window, the second-layer attribution delay processing structure continuously compares the recurrence probability of the field status summary under the same power facility, the same physical location, and similar operating conditions by searching the spatial database. The calculation logic for the recurrence probability is the total number of times the field status summary appears within the observation period divided by the total number of scans within the delay attribution window. If the value is less than 30%, the second-layer attribution delay processing structure determines the phenomenon as an occasional transient interference and executes the appropriate procedure. The attribution clearing operation is performed. If the recurrence probability consistently exceeds 75%, the second-layer attribution delay processing structure initiates feature extraction based on cross-correlation coefficients to verify the matching degree between the current feature sequence and the preset material degradation sample library. When the matching degree reaches 0.9 or higher, the second-layer attribution delay processing structure can determine the corresponding attribution state and trigger the corresponding material degradation maintenance chain. During the entire duration of the delayed attribution window, the gating freeze execution module prohibits modification of the initial feature vector of the field state summary to ensure data consistency during the comparison process. If, at the end of 48 hours, the recurrence probability is in the intermediate range between 30% and 75%, the second-layer attribution delay processing structure must automatically extend the delayed attribution window by 24 hours and simultaneously request the state summary extraction module to increase the sampling density until the system obtains a clear attribution conclusion.

[0017] Preferably, when the gated freeze execution module executes the operation to prohibit consecutive identical attribution, it follows a low-level control specification based on attribution instability priority to prevent the diagnostic results from falling into local optima or pseudo-steady states. If the field status summary was temporarily attributed to environmental factors in the previous observation period, then in the current subsequent observation period, the gated freeze execution module must forcibly block the permission of the field status summary to use environmental conclusions through a logical gate lock. At this time, the gated freeze execution module forcibly switches the state attribute of the power facility to the undefined drift state, withdraws all initially assigned tags, and clears the attribution cache. If the field status summary was attributed to equipment material degradation in the previous period, then in the subsequent period, the use of the equipment material conclusion is also prohibited, and the power facility is disassembled back to the undefined drift state for re-processing. The system employs a mechanism that cuts off the convergence direction of causes, requiring new verification support to be generated in each cycle. Only when the field status summary spans more than three environmental change cycles and appears synchronously under different load levels, satisfying the multi-round confirmation conditions, is the gating freeze execution module allowed to remove the field status summary from the undefined drift state. During the undefined drift state, the system performs a comprehensive feature comparison across cycles, comparing the correlation between historical current fundamental distortion and the current temperature rise rate. If, after four consecutive attribution reset operations, the feature items of the field status summary still do not show a clear fault indication, the gating freeze execution module forcibly places it in the observation queue, strictly prohibiting it from triggering the material degradation maintenance chain, thereby effectively preventing erroneous attribution conclusions due to operating condition coupling.

[0018] Preferably, after receiving the attribution status, the scheduling operation control module embeds it into the underlying logic of the scheduling permission rule as a first-level constraint condition. The scheduling operation control module maintains a real-time equipment status mapping table and classifies the power facilities into stable, undefined drift, frozen, or pending-review states according to diagnostic conclusions. For power facilities marked as frozen, undefined drift, or pending-review states, the scheduling operation control module executes a global exclusion command, completely removing them from the first-priority scheduling pool in the scheduling permission rule. In this state, the scheduling operation control module strictly prohibits assigning any operation tasks exceeding 50% of the rated load to the power facility and deprives it of participation in grid frequency regulation or emergency backup. The power facility is qualified for use; and for the power facility that has been verified through at least 5 observation periods and whose physical performance indicators have always been within 2% of the preset threshold deviation, even though the attribution state has been repeatedly withdrawn, the dispatch operation control module upgrades its reliability level from low to reliable level in the dispatch permission rules; subsequently, the dispatch operation control module reintegrates the power facility as a reliable carrying unit into the operation pool, allowing it to undertake normal load allocation tasks to balance the overall supply and demand gap of the system; when performing allocation, the dispatch operation control module must verify whether there is a potential accelerated deterioration trend of the equipment by querying the state transition data. If the transition probability is found to exceed 0.15, the load stripping operation must be re-executed immediately to ensure that the safety bottom line of power production is not touched.

[0019] Preferably, when applying the capacity balance constraint, the scheduling operation control module establishes a dynamic equipment fatigue sharing mechanism. If a specific type of power facility is frequently selected by the scheduling task due to its long-term stable attribution status, and the cumulative operating time exceeds a preset 200 hours, the scheduling operation control module must automatically initiate a load bearing qualification degradation procedure. Under this procedure, the scheduling operation control module modifies the scheduling instruction set to forcibly distribute the scheduling task originally belonging to the equipment to other power facility units that are under observation and whose physical performance deviation is less than 1%. When a specific power facility remains in the drift freeze layer due to unstable attribution, the scheduling operation control module prohibits issuing abandonment orders. Instead of issuing repair or long-term shutdown orders, the system precisely categorizes these devices into the conservative operation queue. Devices in this queue are subject to extremely stringent operational restrictions, stipulating that their instantaneous load must not exceed 80% of their rated capacity (the first impact threshold), and that fluctuations in output voltage or frequency must be limited to within 5% of their rated value (the first fluctuation threshold). Furthermore, the scheduling operation control module requires that the tasks undertaken by such devices must be completed within 15 minutes. During the execution of these restricted tasks, the gating freeze execution module must continue to collect the on-site status summary at double the sampling frequency and transmit it in real-time to the feedback parameter correction module to ensure that the material degradation process under the nursing operation is under continuous and controlled monitoring.

[0020] Preferably, the feedback parameter correction module acts as the system's self-evolution center, performing logical verification by correlating the load-bearing results with the state transition data. If a specific on-site state summary is detected to have spontaneously subsided within 72 hours of being pressed into the drift-freezing layer, or if cross-verification of infrared imaging data and electrical signal data determines that the signal is not related to the physical material damage of the power facility, then the feedback parameter correction module determines that a false alarm interference has occurred. In this case, the feedback parameter correction module forcibly captures the event as a control failure sample and generates a premature stabilization tendency label accordingly. The feedback parameter correction module performs multi-dimensional feature vector mapping on the premature stabilization tendency label and the current operating environment parameters such as humidity, ambient temperature, and harmonic content. Binding storage; subsequently, the feedback parameter correction module executes a reverse feedback instruction, directly acting on the gating freeze execution module, forcibly raising the judgment threshold under the same operating environment parameters in the next control cycle from 0.85 to 0.95, while simultaneously increasing the value of the preset threshold by 20%; through this numerical correction based on historical error experience, the system cuts off the logical path at the underlying level from the possibility of physical drift of a certain magnitude being incorrectly locked again; after completing the parameter distribution, the feedback parameter correction module must monitor the interception rate change of the gating freeze execution module in the next 24 hours. If the interception rate increase is lower than expected, the feedback parameter correction module will trigger a secondary correction constraint until the frequency of the control failure sample drops below the preset safety threshold of one ten-thousandth.

[0021] Preferably, when performing diagnostic efficiency optimization, the feedback parameter correction module focuses on identifying and correcting the delay control defect. If, during a specific period of delay observation by the gated freeze execution module, the power facility, after experiencing a normal load impact from the dispatch operation control module, records a sudden drop in insulation strength exceeding 15% within 4 hours in its state transition data, the system determines that the power facility has entered a critical state of material degradation. Because the diagnosis in this case lags behind the degradation rate of the physical entity, the feedback parameter correction module forcibly defines the process as the delay control defect and simultaneously generates a delay attribution insufficiency tag. The feedback parameter correction module transmits the tag to the gated freeze execution module in real time via a dedicated feedback link. The system, in conjunction with the scheduling and operation control module, upon receiving the "delay attribution insufficiency" tag, must forcibly reduce the duration of the delay attribution window from 48 hours to 12 hours when facing the same operating conditions, thereby compressing the overall observation time when similar anomalies appear. Simultaneously, the scheduling and operation control module must synchronously increase the execution priority of the state to be reviewed, isolating it as a high-risk unit. During this process, the feedback parameter correction module will also trigger a joint review operation on the control boundary and the judgment threshold. By analyzing the weak precursor signals before the critical state, the sensitivity of the diagnostic model is adjusted incrementally with a step size of 0.02 to ensure that the system can achieve closed-loop response in the very early stage of degradation.

[0022] Preferably, when performing the final control boundary correction, the feedback parameter correction module introduces rigorous decision damping and cross-confirmation operations to ensure system robustness. The specification requires that any modification to the judgment parameters and control thresholds must be verified through at least two independent scenarios with highly consistent meteorological characteristics and load curves. The feedback parameter correction module is only permitted to write parameters when the correction directions of the two independent verifications are consistent and the confidence level exceeds 95%. During parameter updates, the feedback parameter correction module strictly limits the step size of a single adjustment, adopting a step-by-step limited adjustment mode, stipulating that the correction magnitude of a single adjustment to the judgment threshold or the preset threshold must not exceed 5% of the original baseline value. The underlying mechanism of the decision damping operation introduces a first-order inertial filtering stage to prevent the control boundary from being affected by the first-order inertial filter. The long-term period refers to directional oscillations caused by instantaneous operating condition fluctuations or single random overshoot events within 30 minutes. This damping effect makes the changes in system parameters exhibit smooth exponential convergence characteristics, thus enabling the entire diagnostic architecture to remain stably in a preset equilibrium state during the second long-term period, i.e., long-term operation of more than 7 days. After the second long-term period ends, the feedback parameter correction module performs a global consistency evaluation on all state transition paths during the period. If it is found that the accuracy of the material degradation report output by the system remains above 99%, the current control boundary is defined as the standard operating baseline. Through this technical description system composed of logical mandatory constraints, multi-layer verification support, and damping smooth control, the intelligent diagnostic and energy efficiency optimization system for detecting material degradation of power facilities achieves seamless monitoring and high-precision early warning of the operating status of power facilities.

[0023] (III) Beneficial Effects

[0024] This invention provides an intelligent diagnostic and energy efficiency optimization system for detecting material degradation in power facilities. It offers the following advantages:

[0025] 1. This invention, through the synergistic effect of the gated freeze execution module and the drift freeze layer, can retain and observe the on-site status summary that does not meet the multi-round confirmation conditions, effectively intercepting false degradation conclusions caused by environmental phenomena or instantaneous anomalies; by using the prohibition of continuous identical attribution operations and the second-layer attribution delay processing structure, the system forcibly blocks the blind progression of fault levels, ensuring that the triggering of the material degradation maintenance chain is based on a stable judgment after multi-cycle, cross-operating condition feature comparison.

[0026] 2. This invention relies on the capacity balance constraints and scheduling permission rules in the scheduling operation control module to reasonably classify facilities in the observation period into the conservative operation queue, thereby maximizing the utilization of the remaining load-bearing capacity of the equipment while distributing the risk of material deterioration. In conjunction with the feedback parameter correction module to capture control failure samples and perform decision damping operations, the system can dynamically adjust the judgment threshold and control boundary based on the critical state data after load impact, and use incremental adjustment with variable step size to prevent directional oscillations caused by operating condition fluctuations. Attached Figure Description

[0027] To make the content of this invention easier to understand, the invention will be further described in detail below with reference to specific embodiments and accompanying drawings, wherein:

[0028] Figure 1 This is an overall flowchart of an intelligent diagnostic and energy efficiency optimization system for detecting material degradation in power facilities, as described in this invention. Detailed Implementation

[0029] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0030] This invention provides an intelligent diagnostic and energy efficiency optimization system for detecting material degradation in power facilities, including a system for the operation and maintenance of 500 kV transformers in large substations.

[0031] At the substation power facility operation and maintenance site, the status summary extraction module, as the core of the system's perception layer, continuously acquires the real-time operating parameters of the transformer through current transformers, voltage sensors, and partial discharge detectors deployed on the high-voltage side bushings, neutral grounding wires, and tank surfaces. The analog-to-digital converter inside the status summary extraction module converts the acquired analog electrical signals into a digital sequence with a sampling frequency of 12000 Hz. It then uses a fast Fourier transform algorithm to extract characteristic frequency components and encapsulates the extracted current RMS value, harmonic distortion rate, and insulation dielectric loss tangent value into a data packet representing the on-site status summary. In the preprocessing stage, the status summary extraction module uses a preset min-max normalization algorithm to map the original physical parameters such as current, voltage, and loss tangent value to a per-unit value range between 0 and 1. Specifically, this is achieved by calculating the current value and the minimum value of the measurement range. The difference is calculated by dividing by the difference between the maximum and minimum values ​​of the measurement range, ensuring that power parameters of different dimensions are compared within the same order of magnitude. If the current value collected by the sensor reaches 130% of the sensor's rated full-scale range, the status summary extraction module immediately performs a data truncation operation and triggers an abnormal data marker bit to prevent overflow errors from interfering with subsequent processing. Subsequently, a globally unique timestamp index consisting of 17 bits is added to the header of the data packet, and the field status summary is transmitted to the gated freeze execution module through the internal high-speed parallel bus. The status summary extraction module completes a full data scan cycle every 200 milliseconds. If a valid sensor receipt is not obtained within the scan cycle, the status summary extraction module forcibly fills the field status summary of the cycle with the value of the previous valid cycle and simultaneously sends a warning signal of data quality degradation to the gated freeze execution module.

[0032] The specific process of the system's overall diagnosis and scheduling is as follows: After receiving the field status summary, the gated freeze execution module immediately calls the internal real-time comparator to execute the environmental phenomenon identification protocol, comparing the normalized amplitude feature item in the field status summary with the preset threshold of 0.15 point by point; if the field status summary is the first occurrence in the current observation sequence, and its normalized amplitude is only 0.12, which is lower than the preset threshold of 0.15, the gated freeze execution module determines the field status summary as the environmental phenomenon, performs a mandatory path blocking operation, blocks the calculation path of the field status summary towards any material degradation conclusion, and strictly prohibits such low amplitude fluctuations from being included in the damage accumulation model of the power facility; the gated freeze execution module redirects the environmental phenomenon to the temporary storage area in the drift freeze layer and marks it as the observation data stream by modifying the routing control bit of the data packet, restricts the system from triggering equipment damage determination during the retention of the observation data stream, and forcibly closes the maintenance action prompt interface to prevent false alarms caused by background noise; when the field status Upon entering the unconfirmed deterioration state, the gating freeze execution module pushes the current state summary into the drift freeze layer using a first-in-first-out stack management mechanism and verifies whether it meets the multi-round confirmation conditions. Specifically, it requires a monotonically increasing characteristic over five consecutive sampling periods, with each period's offset increment greater than 0.05, and its cumulative amplitude must be stable above 0.12. When the comprehensive confidence index exceeds the judgment threshold of 0.85, the gating freeze execution module forms the single-time stability judgment and forcibly redirects it to the second-layer attribution delay processing. The structure performs secondary temporal verification, strictly prohibiting direct progression to the first fault level; the second-layer attribution delay processing structure allocates a delay attribution window with a set duration of 48 hours, and continuously compares the recurrence probability of the field status summary at the same power facility and the same physical location by searching the spatial database; if the recurrence probability is less than 30%, the second-layer attribution delay processing structure determines the phenomenon as an occasional transient interference and performs attribution clearing operation; if the recurrence probability stably exceeds 75% and the matching degree between the feature sequence and the preset material degradation sample library reaches 0.92, then the corresponding material degradation maintenance chain is triggered; when performing attribution judgment, the gating freeze execution module follows the processing rule of prioritizing attribution instability. If the field status summary was temporarily attributed to environmental factors in the previous observation period, then in the current subsequent observation period, the gating freeze execution module forcibly blocks the permission of the field status summary to use the environmental conclusion, and forcibly switches the state attribute of the power facility to the undefined drift state; only when the field status summary spans more than 3 environmental change cycles, and appears synchronously under different load levels and meets the multi-round confirmation conditions, is it allowed to be removed from the undefined drift state, otherwise the attribution reset operation continues to be executed; after receiving the attribution state, the scheduling operation control module embeds it as the first-level constraint condition into the underlying logic of the scheduling permission rule, and completely removes the power facilities marked as the frozen state, the undefined drift state, and the pending review state from the first priority scheduling pool by maintaining the real-time equipment state mapping table; the scheduling operation control module strictly prohibits assigning any operation task exceeding 50% of its rated load to such equipment. The system will deprive equipment of its right to participate in power grid frequency regulation. For equipment that has been verified through five observation cycles and whose physical performance deviation is within a preset threshold of 2%, the scheduling operation control module will upgrade its reliability level to the reliable level and reinstate it to the operation pool to undertake load allocation tasks. When the cumulative operating time of a specific type of power facility exceeds a preset 200 hours, the scheduling operation control module will automatically initiate a load carrying qualification downgrade procedure, distributing scheduling tasks to other equipment units under observation with a physical performance deviation of less than 1%. For equipment that remains in the drift freeze layer, the scheduling operation control module will prohibit issuing repair commands and instead classify it into the conservative operation queue, limiting its instantaneous load to no more than 80% of its rated capacity (the first impact threshold) and its output voltage fluctuation range to within 5% of its rated value (the first fluctuation threshold). The task must also meet the condition of rapid exit within 15 minutes. During the execution of the above-mentioned restricted tasks, the gating freeze execution module will continue to collect the field status summary at double the sampling frequency and transmit it to the feedback parameter correction module.

[0033] The feedback parameter correction module, acting as the system's self-evolution center, performs logical verification by correlating the load-bearing results with the state transition data. If a specific on-site state summary is detected to have spontaneously subsided within 72 hours of being pressed into the drift-freezing layer, or if cross-verification of infrared imaging data and electrical signal data determines that the signal is unrelated to the physical damage to the power facility, the feedback parameter correction module determines that a false alarm interference has occurred. In this case, the feedback parameter correction module forcibly captures the event as a control failure sample and generates a premature stabilization tendency label accordingly. The feedback parameter correction module correlates the premature stabilization tendency label with current humidity, ambient temperature, and harmonic content, etc. The operating environment parameters are bound and stored as multi-dimensional feature vectors. Subsequently, the feedback parameter correction module executes a reverse feedback instruction, which directly acts on the gating freeze execution module, forcibly raising the judgment threshold under the same operating environment parameters in the next control cycle from 0.85 to 0.95, while simultaneously increasing the value of the preset threshold by 20%. Through this numerical correction based on historical error experience, the system cuts off the processing path of physical drift of a certain magnitude being erroneously locked again at the underlying level. After completing the parameter distribution, the feedback parameter correction module monitors the interception rate change of the gating freeze execution module in the next 24 hours. If the interception rate increase is lower than expected, the feedback parameter correction module will trigger a secondary correction contract.

[0034] Furthermore, when performing diagnostic efficiency optimization, the feedback parameter correction module focuses on identifying and correcting the delay control defect. If, during a specific period of delay observation, the power facility, after experiencing a normal load impact from the dispatch operation control module, records a sudden drop in insulation strength exceeding 15% within 4 hours in its state transition data, the system determines that the power facility has entered a critical state of material degradation. Because the diagnosis in this case lags behind the degradation rate of the physical entity, the feedback parameter correction module forcibly defines this process as a delay control defect and simultaneously generates a delay attribution insufficiency label. The feedback parameter correction module uses a dedicated feedback chain... The system transmits the tag to the gating freeze execution module and the scheduling operation control module in real time. Upon receiving the insufficient delay tag, the system must forcibly reduce the duration of the delay assignment window from 48 hours to 12 hours when facing the same working conditions, thereby compressing the overall observation time when similar anomalies appear. At the same time, the scheduling operation control module must synchronously increase the execution priority of the state to be reviewed, treating it as a high-risk unit for key isolation. During this process, the feedback parameter correction module will also trigger a joint review operation on the control boundary and judgment threshold, and adjust the sensitivity of the diagnostic model by an incremental step of 0.02 by analyzing the weak precursor signals before the critical state.

[0035] When performing the final control boundary correction, the feedback parameter correction module introduces rigorous decision damping and cross-confirmation operations to ensure system robustness. The specification requires that any modification to the judgment parameters and control thresholds must be verified through at least two independent scenarios with highly consistent meteorological characteristics and load curves. The feedback parameter correction module is only permitted to write parameters when the correction directions of the two independent verifications are consistent and the confidence level exceeds 95%. During parameter updates, the feedback parameter correction module strictly limits the step size of each adjustment, adopting a step-by-step limited adjustment mode, stipulating that the correction magnitude of a single judgment threshold or preset threshold must not exceed 5% of the original baseline value. The underlying mechanism of the decision damping operation introduces a first-order inertial filter to block directional oscillations in the control boundary caused by instantaneous operating condition fluctuations or single random overshoot events within the first time period (30 minutes). This damping effect makes the changes in system parameters exhibit smooth exponential convergence characteristics, thereby enabling the entire diagnostic architecture to remain stably in a preset equilibrium state during the second time period (more than 7 days of long-term operation). After the second duration period ends, the feedback parameter correction module performs a global consistency assessment on all state transition paths during the period. If the accuracy of the material degradation report output by the system is found to be above 99%, the current control boundary is defined as the standard operating baseline. Through this technical system consisting of mandatory constraints of processing specifications, multi-layer verification support, and damping smooth control, the intelligent diagnosis and energy efficiency optimization system for detecting material degradation of power facilities achieves seamless monitoring and high-precision early warning of the operating status of power facilities.

[0036] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An intelligent diagnostic and energy efficiency optimization system for detecting material degradation in power facilities, characterized in that, include: The module includes a status summary extraction module, a gating freeze execution module, a scheduling operation control module, and a feedback parameter correction module. The status summary extraction module obtains the on-site status summary of the power facilities, and transmits the on-site status summary to the gated freeze execution module after performing feature normalization processing. The gated freeze execution module receives the field status summary and determines whether the field status summary meets the multi-round confirmation conditions based on the working conditions. If the field status summary does not meet the multi-round confirmation conditions, the field status summary is classified into an unconfirmable deterioration state and pushed into the drift freeze layer to retain the observation data stream. When the site status summary meets the multi-round confirmation conditions, a delayed attribution window is allocated to the site status summary and the second-level attribution delay processing structure is entered. At the same time, the operation of prohibiting continuous identical attribution is executed to form a single stable judgment, thereby determining the corresponding attribution state and triggering the corresponding material deterioration maintenance chain. The attribution state is then transmitted to the scheduling operation control module. The scheduling and operation control module matches the corresponding scheduling permission rules and capacity balance constraints according to the attribution state, divides the power facilities into stable state, undefined drift state, frozen state or pending review state, and then assigns them to the corresponding operation pool or conservative operation queue to allocate scheduling tasks, and transmits the load carrying results and state transition data to the feedback parameter correction module. The feedback parameter correction module receives the load carrying results and the state transition data. It captures control failure samples by analyzing the self-fading of the field state summary, or determines delay control defects by analyzing the critical state of the power facilities after being subjected to the scheduling task. Based on this, it generates premature stability tendency labels or delay attribution insufficient labels, and adjusts the preset thresholds and judgment thresholds in the gating freeze execution module and the control boundaries in the scheduling operation control module by incremental adjustment of the step size through decision damping operation and cross-confirmation operation.

2. The intelligent diagnostic and energy efficiency optimization system for detecting material degradation in power facilities according to claim 1, characterized in that: When the gated freeze execution module controls the unconfirmed deterioration state and the drift freeze layer, it intercepts the first occurrence of the field state summary with an amplitude lower than the preset threshold as an environmental phenomenon, blocking the path that the field state summary leads to the conclusion of material deterioration; it retains the observation data stream of the environmental phenomenon in the drift freeze layer, and restricts the system from triggering the corresponding equipment body damage judgment and maintenance action prompt during the retention of the observation data stream; After the field status summary undergoes multiple cyclic filtering and forms the single stability judgment, the gating freeze execution module blocks the path of the single stability judgment to the first fault level and forces the field status summary to be sent into the second layer of attribution delay processing structure. During the opening of the delay attribution window, the recurrence probability of the field status summary on the same device and the same location is continuously compared.

3. The intelligent diagnostic and energy efficiency optimization system for detecting material degradation in power facilities according to claim 1, characterized in that: When the gating freeze execution module performs the operation to prohibit continuous identical attribution, if the field status summary was temporarily attributed to environmental factors in the previous cycle, then in the next cycle, the permission for the field status summary to use the environmental factors is blocked, and the power facility is switched to the undefined drift state for re-observation; if the field status summary was temporarily attributed to the equipment body in the previous cycle, then in the next cycle, the conclusion of using the equipment body is prohibited, and the power facility is dismantled back to the undefined drift state. By cutting off the cause convergence direction and maintaining the priority of attribution instability, the system is prevented from forming the material degradation maintenance chain prematurely; only when the field status summary appears synchronously under adjacent operating conditions and meets the multi-round confirmation conditions is the field status summary allowed to move out of the undefined drift state. Otherwise, the attribution reset operation continues to be performed, and the field status summary is kept in the undefined drift state for comprehensive and cross-cycle feature comparison repeated observation.

4. The intelligent diagnostic and energy efficiency optimization system for detecting material degradation in power facilities according to claim 1, characterized in that: When the scheduling operation control module matches the scheduling permission rules based on the attribution status, it uses the attribution status as the first-level constraint condition for determining whether the power facility is allowed to undertake the scheduling task. For power facilities in the frozen state, the undefined drift state, and the pending review state, the scheduling operation control module excludes them all from the first priority scheduling pool in the scheduling permission rules, depriving them of their qualification to undertake rated load operation tasks. For power facilities that have undergone multiple observation periods and whose attribution status has been repeatedly withdrawn or reset but whose operation status has not deviated from the preset threshold range, the scheduling operation control module upgrades their reliability level in the scheduling permission rules and includes them as reliable carrying units in the operation pool, thereby balancing safety risks.

5. The intelligent diagnostic and energy efficiency optimization system for detecting material degradation in power facilities according to claim 1, characterized in that: When the scheduling operation control module applies the capacity balance constraint, if a specific type of power facility is frequently selected due to stable attribution status, its load-bearing qualification is automatically downgraded, and subsequent scheduling tasks are forcibly distributed to other power facility units that are under observation and have not experienced physical performance deviations from the preset range. When a power facility is continuously in the drift freeze layer and the attribution cannot be stable, the scheduling operation control module prohibits the execution of the abandoned repair action, but directly moves it into the conservative operation queue, limiting it to only undertake specific scheduling tasks within the first impact threshold and the first fluctuation threshold, and which can be quickly exited. By limiting the output intensity of the scheduling task, the risk of material degradation of the power facility is distributed. At the same time, the gated freeze execution module continues to collect the field status summary under the preset load operation state and transmits it to the feedback parameter correction module to perform uninterrupted material degradation evolution status data tracking.

6. The intelligent diagnostic and energy efficiency optimization system for detecting material degradation in power facilities according to claim 1, characterized in that: When the feedback parameter correction module processes the load bearing results and the state transition data, if a specific field state summary is pressed into the drift freeze layer and then the field state summary is detected to have spontaneously faded in subsequent observation periods, or if the field state summary is determined to be unrelated to the actual material degradation of the power facilities through cross-verification of multi-source data, then the feedback parameter correction module will capture the spontaneously fading event as a control failure sample, generate a premature stabilization tendency label, and associate and bind it with the current operating environment parameters. Subsequently, the premature stabilization tendency label is fed back to the gating freeze execution module, thereby forcibly and directly increasing the values ​​of the judgment threshold and the preset threshold under the same operating environment conditions in the next control period, thereby cutting off the evolution path of physical drift being incorrectly locked within the set range.

7. The intelligent diagnostic and energy efficiency optimization system for detecting material degradation in power facilities according to claim 1, characterized in that: When the feedback parameter correction module establishes the delay attribution deficiency label, if during the specific observation period when the gated freeze execution module fails to complete stable attribution, after the power facility has experienced the load impact of the scheduling task, the state transition data clearly reveals that the power facility itself has entered the critical state, then the system will define the state as the delay control defect and generate the delay attribution deficiency label accordingly. The feedback parameter correction module transmits the insufficient delay attribution tag to the gating freeze execution module and the scheduling operation control module in real time. Under the same working conditions in the future, it forcibly shortens the set duration of the delay attribution window and simultaneously increases the execution priority of the pending review state. This reduces the overall observation time when similar anomalies are displayed and triggers the review operation for the control boundary and the judgment threshold.

8. The intelligent diagnostic and energy efficiency optimization system for detecting material degradation in power facilities according to claim 1, characterized in that: When the feedback parameter correction module calls the decision damping operation and the cross-confirmation operation to correct the control boundary, it stipulates that the correctness of any correction of the judgment parameter and the control threshold must be verified by at least two independent scenarios with consistent environmental characteristics. After the scenario verification is satisfied, the feedback parameter correction module strictly limits the step size of the single parameter adjustment, and adopts a step-by-step limited adjustment mode to make directional adjustments to the judgment threshold, the preset threshold, the setting duration of the delay attribution window in the gating freeze execution module, and the scheduling permission rules in the scheduling operation control module. The decision damping operation is used to prevent the control boundary and the judgment threshold from directional oscillations caused by instantaneous operating condition fluctuations or single sporadic events within the first time period, thereby making the system stably maintain a preset equilibrium state within the second time period.

Images