Search

CN-122017663-A - No-load abnormality monitoring method and early warning system for aircraft power supply system

CN122017663ACN 122017663 ACN122017663 ACN 122017663ACN-122017663-A

Abstract

The invention discloses an empty load abnormality monitoring method and an early warning system of an aircraft power supply system, which relate to the technical field of electric variable measurement and comprise the steps of executing an empty load exclusive reference calibration after confirming that the system enters a stable empty load working condition, and establishing an empty load operation reference database; the method comprises the steps of acquiring idle electricity variable and environmental parameter by a multichannel unit calibrated synchronously at nanosecond level, extracting abnormal characteristics by means of time domain and frequency domain joint analysis, correcting data by means of an idle exclusive environment-aging coupling compensation model, quantifying risks by means of an idle exclusive risk scoring formula, judging abnormal types and grades by means of an idle exclusive abnormality identification model, completing fault tracing by means of a joint positioning model, triggering grading early warning and encrypting stored data. The method can accurately capture the no-load weak abnormal signal, improve the early fault recognition rate and ensure the operation safety of the aircraft power supply system.

Inventors

  • YAN XIAOLONG

Assignees

  • 陕西星辰电子技术有限责任公司

Dates

Publication Date
20260512
Application Date
20260409

Claims (10)

  1. 1. The method for monitoring the no-load abnormality of the aircraft power supply system is characterized by comprising the following steps of: After a working condition switching signal of an aircraft power supply system is received and a stable no-load operation working condition is entered, an no-load exclusive reference parameter calibration operation is executed, phase voltage, line voltage, no-load leakage current, output frequency, harmonic total distortion rate, phase sequence phase difference, output impedance and ripple coefficient electric variable reference data under a rated no-load working condition are collected, and an no-load normal operation reference database is built; The system comprises a multichannel synchronous acquisition unit, a system reference clock and a system reference clock, wherein nanosecond synchronous calibration is carried out on each acquisition channel, so that deviation of phase sequence phase difference and impedance calculation sensitivity to time delay under no-load is eliminated; redundant dual-channel parallel acquisition is adopted for the no-load leakage current weak signal channel, the validity of data is verified, and the environmental temperature, humidity and atmospheric pressure parameters are synchronously acquired; Performing time domain and frequency domain joint analysis on the acquired real-time electrical variable data, extracting voltage fluctuation, current mutation, frequency offset and impedance change characteristics from the time domain by adopting a sliding window method, extracting harmonic distribution characteristics from the frequency domain by fast Fourier transformation to obtain an abnormal characteristic parameter set, and finishing normalization processing and invalid characteristic elimination; Based on the environment parameters and the equipment no-load aging state, adopting an no-load dedicated environment-aging coupling interference compensation model to correct, wherein a conventional electric variable adopts independent environment compensation, no-load leakage current is introduced into an aging-environment coupling correction term, a temperature interference coefficient is dynamically adjusted along with the equipment no-load aging degree, the covering effect of coupling interference on the weak leakage current characteristic is quantized, and the compensation correction of the electric variable and the characteristic parameter is completed; Comparing the corrected characteristic parameters with a reference threshold value, calculating an abnormal risk comprehensive score by adopting an idle load exclusive risk score formula, and amplifying the risk weight of idle load weak deviation characteristics; Combining a system topological structure, adopting a Bayesian probability reasoning and idle fault feature matching combined positioning model, correcting posterior probability through feature matching degree, solving positioning deviation caused by insufficient idle fault features, and finishing abnormal module positioning and fault tracing analysis; Triggering the early warning signals of the corresponding levels according to the determined abnormal risk levels, and synchronously encrypting and storing the full stream data in a closed loop.
  2. 2. The aircraft power system no-load anomaly monitoring method of claim 1, wherein the anomaly risk composite score is calculated by the formula: Wherein, the method comprises the steps of, The empty abnormal risk comprehensive score of the aircraft power supply system is given, To assist in the total number of core electrical variable dimensions evaluated, Is the first The weight coefficients of the dimensions of the individual electrical variables, Is the first The real-time acquisition of the values of the individual electrical variables, Is the first The reference calibration value of the individual electrical variable, Is the first The maximum allowable limit of the individual electrical variables, Is the first The nonlinear correction index of the individual electrical variables, For the total number of characteristic frequencies for harmonic analysis, Is the first The impact weight of the individual characteristic frequency harmonics, Is the first The difference between the real-time distortion rate of the individual characteristic frequency harmonics and the reference distortion rate, Is the first Maximum allowable distortion difference of individual characteristic frequency harmonics, For the anomaly duration to affect the coefficient, For the duration of the deviation of the anomaly characteristic parameter from the reference threshold.
  3. 3. The method for monitoring the no-load abnormality of the aircraft power supply system according to claim 1, further comprising an idling condition reference threshold adaptive updating step of collecting full-life-cycle idling operation data, environment parameters and idling aging characteristic parameters in a fixed period, performing adaptive iterative updating on the reference threshold by combining a historical abnormality recognition result with erroneous judgment and omission judgment feedback, and synchronously updating a dynamic coefficient table of a coupling compensation model, wherein the updated parameters are used for subsequent abnormality monitoring.
  4. 4. The method for monitoring the no-load anomaly of the aircraft power supply system according to claim 1, wherein the calculation formula of the anomaly occurrence posterior probability of the anomaly module positioning and fault tracing analysis is as follows: ; To present abnormal characteristic set When an abnormality occurs at the first Power supply module Is used to determine the posterior probability of (1), For the total number of topology modules of the aircraft power supply system, Is the first When the power supply modules are abnormal, abnormal feature sets appear Is a function of the conditional probability of (1), Is the first The prior failure probability of the individual power supply modules, Is an abnormal feature set And the first The number of matching features of the fault signature library of the individual power modules, Is an abnormal feature set Is a total feature quantity of the (c).
  5. 5. The method for monitoring the empty load abnormality of the aircraft power supply system according to claim 1 is characterized in that the empty load exclusive abnormality identification model adopts a serial fusion architecture of a multi-scale convolutional neural network and a two-way long-short-period memory network, the front end extracts space dimension features, the rear end extracts time sequence dimension features, an abnormality classification result is output through a full-connection layer, the model adopts full-life-period empty load normal data, empty load abnormal simulation data and historical empty load actual measurement data, leakage current samples in different aging stages and environments are additionally added to construct a training set, cross entropy loss is used as an optimization target, and the training is completed through a self-adaptive moment estimation algorithm, so that incremental learning optimization is supported.
  6. 6. The method for monitoring the no-load abnormality of the aircraft power supply system according to claim 1, wherein the fitting process of the aging-environment coupling correction term is that equipment no-load aging stage is divided into a plurality of sections, interference coefficients of humiture on no-load leakage current are fitted for each section respectively to form a dynamic coupling coefficient table, and the interference coefficients of the corresponding sections are called to execute compensation according to the current equipment aging state during compensation.
  7. 7. The method for monitoring the empty load abnormality of the aircraft power supply system according to claim 1, wherein the verification and switching rule of the redundant dual-channel acquisition architecture is that two sets of same-precision sensing units of an empty load leakage current acquisition channel synchronously output data, two sets of data deviation values are calculated, and when the deviation exceeds a preset limit value, the redundant dual-channel acquisition architecture is automatically switched to a standby channel and channel fault early warning is triggered.
  8. 8. An aircraft power supply system no-load abnormality early warning system according to any one of claims 1 to 7, characterized by comprising a multi-dimensional electric variable acquisition module, a reference data management module, an abnormality feature extraction module, an environment interference compensation module, an intelligent abnormality identification module, an abnormality positioning and tracing module, an early warning output module and a full-flow data storage module which are sequentially connected in a communication manner; The multi-dimensional electric variable acquisition module is internally provided with a nanosecond timing synchronization calibration unit, and is used for configuring a redundant dual-channel acquisition architecture aiming at an idle leakage current weak signal channel and completing synchronous acquisition of electric variables and environmental parameters under an idle working condition; The reference data management module is used for executing reference parameter calibration of the no-load working condition of the aircraft power supply system, establishing and maintaining an no-load normal operation reference database exclusive to no-load, supporting execution of self-adaptive iterative update on an no-load reference threshold value and a coupling compensation coefficient based on the no-load operation data of the whole life cycle of the power supply system, the no-load aging characteristic of equipment and the environmental parameter change, and supporting traceability query of historical reference data; The abnormal characteristic extraction module is used for performing time domain and frequency domain joint analysis on the acquired real-time electrical variable data and extracting core abnormal characteristic parameters under corresponding no-load working conditions; The environment interference compensation module is internally provided with a pre-fitted no-load exclusive environment-aging coupling interference compensation model, a dynamic coupling correction unit is configured aiming at no-load leakage current, environment parameters and equipment aging states which are synchronously collected are taken as input, interference amounts of coupling interference on empty load weak current variable data and weak characteristic parameters are quantized, and compensation correction is carried out on real-time electric variable data and abnormal characteristic parameters; The intelligent anomaly identification module is internally provided with a pre-trained no-load exclusive anomaly identification model and a risk quantification evaluation unit, and is used for combining real-time anomaly characteristic parameters with reference database data to identify no-load anomaly types; The anomaly locating and tracing module is internally provided with a combined anomaly locating model based on Bayesian probability reasoning and no-load fault feature matching, calculates the posterior probability of anomaly occurrence of each power module by combining with the aircraft power system topological structure, and executes anomaly module locating and fault tracing analysis; The early warning output module is used for triggering an acousto-optic early warning signal of a corresponding level according to the determined abnormal risk level and synchronously pushing early warning information to the aircraft avionics system and the ground operation and maintenance platform; the full-flow data storage module is used for executing encryption storage on the electrical variable data, the characteristic parameters, the abnormal identification result and the traceability information acquired in the full-flow process, and establishing an operation audit log.
  9. 9. The empty load abnormality early warning system of the aircraft power supply system according to claim 8, wherein the redundant dual-channel acquisition architecture of the multi-dimensional electric variable acquisition module is provided with a built-in data fusion check unit, the built-in data fusion check unit is used for comparing empty load weak signal acquisition data of two sets of sensing units in real time, when two sets of data deviation exceeds a preset limit value, the system is automatically switched to a standby acquisition channel and triggers channel fault early warning, the sampling frequency of the multi-channel synchronous sampling unit is up to 1MHz, and the synchronous error of nanosecond time sequence synchronous calibration is not more than 100ns.
  10. 10. The aircraft power supply system no-load abnormality early warning system according to claim 8, further comprising an airborne edge calculation unit and a ground operation and maintenance platform linkage unit, wherein the airborne edge calculation unit is in communication connection with a multi-dimensional electric variable acquisition module, an intelligent abnormality recognition module, an early warning output module and a full-flow data storage module to complete real-time acquisition, feature extraction, abnormality recognition and local early warning of no-load electric variable data, and the ground operation and maintenance platform receives full-load monitoring data uploaded by the system through an airborne communication link to support centralized monitoring of multiple times of no-load operation data of the aircraft power supply system, statistical analysis of historical no-load fault data, centralized training and optimization updating of an no-load abnormality recognition model and a coupling compensation model, and can generate an operation and maintenance scheme based on the uploaded abnormal data.

Description

No-load abnormality monitoring method and early warning system for aircraft power supply system Technical Field The invention relates to the technical field of electric variable measurement, in particular to an empty load abnormality monitoring method and an early warning system for an aircraft power supply system. Background The aircraft power supply system is a core component of an aircraft avionics system, and the running stability of the aircraft power supply system directly determines the overall power supply safety and the flight safety of the aircraft. The no-load operation working condition is a key operation state in the processes of ground overhaul, pre-check before flight and air working condition switching of the aircraft power supply system, and the electric operation characteristics of the power supply system under the no-load working condition can directly reflect the whole health level and potential fault hidden trouble of the system, so that the no-load operation working condition is a key window period for identifying electric faults in advance and avoiding power supply interruption risks in the process of carrying and operating in a belt. The current mainstream abnormality monitoring technology of the aircraft power supply system focuses on fault monitoring and protection under the belt running working condition, and obvious defects exist in the special monitoring technology aiming at the no-load working condition, the change amplitude of the electrical characteristics of the system under the no-load working condition is far smaller than that of the belt running working condition, the conventional threshold monitoring method is difficult to capture weak abnormal signals, early hidden fault early identification cannot be completed in the no-load stage, fault missed judgment is easy to cause, and fault upgrading occurs when the power supply system is in belt running. The existing monitoring method for the aircraft power supply system mainly has the following core defects: Firstly, the monitoring scheme has no-load working condition exclusive suitability, and the general threshold value and the monitoring logic of the load working condition are adopted, so that the recognition requirement of the no-load working condition weak fault characteristic cannot be adapted. The existing scheme adopts a fixed threshold comparison mode of a single electric variable, can only identify serious faults exceeding a rated limit value, has serious defects of early abnormality and hidden fault identification capability under an empty load working condition, and is easy to be misjudged and missed to judge due to fluctuation interference of the working condition; And secondly, coupling interference between the environment and equipment aging cannot be effectively eliminated. Environmental parameters such as temperature and humidity, atmospheric pressure and the like in an aviation scene have wide variation range, the full life cycle aging of equipment can cause electrical characteristic drift, the prior art only adopts general fixed coefficient environmental compensation, no consideration is given to the influence of the aging and the environment on no-load weak signals, the weak fault characteristics are easily hidden by interference, and the monitoring precision is continuously reduced in long-term operation; Thirdly, the capability of abnormal quantitative evaluation and accurate positioning aiming at no-load working conditions is lacking. Meanwhile, because the fault characteristic quantity is small under the no-load working condition, the conventional fault positioning method is easy to generate positioning deviation, cannot finish the accurate positioning and fault tracing of an abnormal module, and is difficult to provide effective data support for subsequent operation and maintenance; Fourth, no-load condition full-flow closed-loop monitoring system is formed. The prior art does not construct a full-flow special monitoring system from reference calibration, data acquisition, interference compensation and feature recognition to positioning and early warning aiming at no-load working conditions, cannot adapt to the operation requirements of the aviation field on the high reliability and the high safety of a power supply system, and is difficult to meet the health management and control requirements of the full life cycle of an aircraft power supply system. Disclosure of Invention The invention provides an empty load abnormality monitoring method and an early warning system for an aircraft power supply system, which are used for solving the problems in the prior art. In order to achieve the purpose, the invention adopts the following technical scheme that the method for monitoring the no-load abnormality of the aircraft power supply system comprises the following steps: After a working condition switching signal of an aircraft power supply system is received and a stable no-load operation working condition