A Method and System for Equipment Health Assessment and Remaining Life Prediction Based on a Dual Degradation Mechanism

Abstract

This invention discloses a method and system for assessing equipment health and predicting remaining lifespan based on a dual degradation mechanism. The method includes: initializing the equipment's health and vulnerability; updating the health based on a preset natural wear rate within each discrete time step; if a fault event occurs within the current time step, calculating the damage amount based on the severity, duration, and current vulnerability of the fault event, and updating the health and vulnerability accordingly; if the fault is resolved, calculating the recovery amount of the health based on a preset recovery factor, and updating the health accordingly; updating the vulnerability based on a preset forgetting factor; and calculating the remaining lifespan of the equipment based on the updated health and a preset failure threshold. Compared with existing technologies, this invention achieves real-time accurate assessment of equipment health status and precise prediction of remaining lifespan through a dual degradation mechanism of natural aging and fault impact, and its damage memory effect.

Description

Technical Field

[0001] This invention relates to the field of predictive maintenance technology, and in particular to a method and system for assessing equipment health and predicting remaining life based on a dual degradation mechanism. Background Technology

[0002] Traditional maintenance strategies for power facilities often rely on periodic inspections or threshold judgments based on single performance parameters, making it difficult to achieve a comprehensive and dynamic assessment of the real-time operating status of equipment. In actual operation, equipment undergoes gradual natural aging due to long-term use and may also be impacted by sudden random failures. Existing technologies often fail to effectively consider and quantify these two different degradation mechanisms, neglecting the cumulative damage effects of failure events and the performance recovery capability of equipment after repair. This leads to inaccurate assessments of equipment health status and limits the reliability of remaining service life predictions.

[0003] A search revealed that application publication number CN112836379A discloses a method for health diagnosis of equipment in intelligent manufacturing systems. This method establishes a Gauss-Poisson stochastic process model of equipment performance degradation, calculates the ratio of historical cumulative degradation to initial performance as a vulnerability index, and divides this index into different intervals to correspond to discrete health states such as normal and faulty states, thereby achieving vulnerability-based equipment condition monitoring and classification early warning. However, the model constructed by this method focuses on statistical fitting and state mapping of historical performance data. Its evaluation results are essentially static indicators based on cumulative amounts, failing to further distinguish the inherent differences and interactive effects between natural aging and sudden failure mechanisms in the equipment degradation process. Furthermore, it lacks a quantitative description of the continuous evolution of equipment condition and the dynamic dissipation of damage memory.

[0004] Therefore, how to balance the dual degradation effects of natural wear and tear and fault disturbances, and take into account the cumulative damage memory of faults and the equipment's self-repair capabilities, in order to comprehensively and accurately assess the real-time health status of the equipment and accurately predict its remaining lifespan, is a technical problem that needs to be solved. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of the prior art by providing a method and system for assessing equipment health and predicting remaining life based on a dual degradation mechanism.

[0006] The objective of this invention can be achieved through the following technical solutions: According to a first aspect of the present invention, a method for assessing equipment health and predicting remaining life based on a dual degradation mechanism is provided, comprising: Initialize the device's state variables, which include health HI, used to characterize the device's real-time functional state, and vulnerability V, used to characterize the device's accumulated damage memory. Within each discrete time step: The health status HI is updated based on a preset natural attrition rate; If a fault event occurs within the current time step, the damage amount D is calculated based on the severity and duration of the fault event and the current vulnerability V, and the health HI is updated based on the damage amount D. At the same time, the damage amount is added to the vulnerability V. If the fault is resolved, calculate the recovery amount R of health based on the preset recovery factor, and update the health HI based on the recovery amount R; The vulnerability V is updated by decay based on a preset forgetting factor; Based on the updated health status HI and the preset failure threshold, the remaining useful life RUL of the device is calculated.

[0007] As a preferred technical solution, the damage amount D includes the instantaneous impact damage at the time of the fault event and the cumulative damage during the duration of the fault.

[0008] As a preferred technical solution, the formula for calculating the instantaneous impact damage is: in, This is instantaneous impact damage. V represents the severity of the current fault, and V represents the vulnerability at the time the fault occurs.

[0009] As a preferred technical solution, the cumulative damage is calculated based on the fault duration using a logarithmic growth model or a linear capping model.

[0010] As a preferred technical solution, the formula for calculating the recovery amount R is: ,in, The recovery factor corresponding to the current fault type. and These represent the instantaneous impact damage and the continuous cumulative damage caused by the fault, respectively.

[0011] As a preferred technical solution, the decay update process of the vulnerability V is an exponential decay process, specifically including: The damage accumulated from the fault events within the current time step is added to the vulnerability V; The vulnerability V, which is superimposed with the damage amount, is exponentially calculated with the preset forgetting factor γ, and the result is used as the final vulnerability value at the end of the current time step, where γ is a constant between 0 and 1.

[0012] As a preferred technical solution, the health status HI is updated based on the natural depletion rate α, specifically as follows: .

[0013] As a preferred technical solution, the remaining service life RUL The calculation formula is: ,in This is the failure threshold.

[0014] According to a second aspect of the present invention, a device remaining life prediction system for implementing the method is provided, the system comprising: The parameter configuration module is used to load and manage system parameters, including natural wear rate, forgetting factor, failure threshold and parameters of various fault types, and provides parameter query interface for other modules. The status maintenance module, at each time step, stores and updates the device's health HI and vulnerability V based on the natural wear and tear rate and forgetting factor. The event processing engine responds to fault occurrence and resolution events, triggers and executes fault damage update and fault recovery update logic, and drives the state maintenance module to perform corresponding health HI and vulnerability V update calculations. The lifespan prediction module calculates the remaining lifespan (RUL) based on the health status (HI) in the status maintenance module.

[0015] As a preferred technical solution, the system further includes a simulation module, which drives the system to operate according to a preset fault event sequence and parameters, and outputs simulation data and curves showing the changes in health (HI), vulnerability (V), and remaining lifetime (RUL) over time.

[0016] Compared with the prior art, the present invention has the following advantages: 1. This invention achieves a comprehensive and accurate real-time assessment of equipment health status by simultaneously modeling the dual degradation mechanism of natural wear and tear and sudden failures, and quantifying the cumulative damage memory of failures and equipment repair capabilities.

[0017] 2. When calculating the damage amount of a fault, this invention dynamically couples the severity and duration of the fault with the current vulnerability, so that the damage assessment depends not only on the fault itself, but also on the changes in the shock resistance caused by the cumulative damage history of the equipment, thereby improving the accuracy and physical authenticity of the condition assessment.

[0018] 3. This invention precisely distinguishes between instantaneous impact damage and continuous cumulative damage from faults, quantifies the repair effect of different faults through recovery factors, and simulates the natural dissipation of vulnerability with forgetting factors, making the evolution of equipment state more in line with actual physical laws.

[0019] 4. This invention, through a discrete-time step update mechanism, can respond to fault events in real time, intuitively present the step changes and recovery process of health status, and perform rolling updates on the remaining service life, providing immediate and intuitive decision support for predictive maintenance. Attached Figure Description

[0020] Figure 1 This is a graph showing the trend of device health index and vulnerability over time in an embodiment of the present invention; Figure 2 This is a graph showing the trend of remaining equipment lifespan as a function of operating days in an embodiment of the present invention. Figure 3 This is a flowchart of the method of the present invention; Detailed Implementation

[0021] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0022] Example 1: like Figure 3 As shown, this invention provides a method for assessing equipment health and predicting remaining life based on a dual degradation mechanism. The method includes: Step S1: Configure model parameters and initialize device state variables, including health HI and vulnerability V; Step S2: Within each discrete time step, execute the following update process sequentially: Step S21: Update the health status HI based on the natural wear and tear rate to simulate the natural aging of the equipment; Step S22: Determine the fault event based on the equipment operation information. If a fault occurs, calculate the damage amount based on the fault severity, duration, and current vulnerability V, update HI accordingly, and accumulate the damage amount to V. If the fault is resolved, calculate the recovery amount based on the recovery factor and update HI. Based on the forgetting factor, the vulnerability V is updated by decay. Step S3: Based on the updated health status HI, combined with the preset failure threshold and natural wear rate, calculate the remaining service life RUL of the equipment.

[0023] The following provides a detailed explanation of each of the above steps.

[0024] Step S101: System Initialization Initialize the model parameters and device states. The model parameters are loaded through a configuration file. Taking a circuit breaker in a power system as an example, its global parameter configuration includes: The natural loss rate α = 0.00000032 (unit: HI points / second) describes the linear rate of performance degradation of a circuit breaker under ideal operating conditions due to wear of mechanical components and aging of materials. The forgetting factor γ = 0.999999 determines the rate at which the damaged memory recorded by the vulnerability V naturally fades over time. Calculations show that this value corresponds to a half-life of about 8 days, which means that the impact of short-term failures will significantly decrease after about a week. The failure threshold L=20. When the health status HI is lower than this value, the circuit breaker is determined to be faulty and needs to be replaced or overhauled.

[0025] The fault parameter configuration predefines various possible fault events and their impact parameters. For example: The basic destructive force (severity) of fault code 1 (undervoltage alarm) is 15.0 HI point; after the fault is resolved, 80% of the damage caused can be recovered (recovery factor 0.8); if the fault continues, the additional damage caused follows a logarithmic growth model with a growth coefficient k of 0.00008; fault code 4 (overtemperature alarm) is more serious, and its continuous damage increases linearly, up to 3 times the basic severity.

[0026] Step S201: Periodic state update The system performs the following updates cyclically with a fixed time step Δt (e.g., 1 second). Assume that the current time step is t, and the previous state is HI(t-1) and V(t-1).

[0027] Step S2011: The impact of simulation time itself on the equipment Natural depletion of health: The natural loss rate α can be calculated based on the expected lifespan of the equipment. For example, if the circuit breaker is required to last for 10 years (approximately 3.1536 × 10⁻⁶), then... 8 If the linear degradation from HI=100 to 0 within (seconds), then α ≈ 100 / 3.1536×10 8 ≈ 3.17×10 -7 HI points / second.

[0028] Step S2012: Process discrete fault events that occur during this cycle. If no fault occurs, then Vulnerability ; If a fault occurs: for example, an undervoltage alarm is triggered at second t (fault type i=1), its severity S1=15.0; (a) Calculate instantaneous impact damage, i.e., the direct damage caused at the instant the failure occurs: ) (b) Calculate cumulative damage over time: If the fault is not resolved immediately, it will cause additional damage during its duration. Select the model based on the fault configuration: Logarithmic type: , The duration of the fault is considered. This means the rate of damage decreases over time, simulating faults with a large initial impact but gradually saturating subsequent effects. Linear capping type: This means that the damage increases linearly over time until a max_multiplier limit is reached. This simulates faults that cause stable damage over time, but whose destructive effects have physical limits.

[0029] Calculate persistent damage, which is the further, cumulative damage to the system that occurs if the fault is not resolved immediately. The formula for persistent cumulative damage is: ,in It is the total time from the occurrence of the fault to its resolution, calculated by subtracting the initial time. The effect precisely isolates any additional damage caused during the duration.

[0030] (c) Update status: The total damage to the health index caused by a complete failure event is: ; Update health: Health decreases due to damage; Update vulnerability: Damage is remembered, increasing future vulnerability; If the fault is resolved: Assuming fault i is resolved at time t, the system will recover a portion of the health lost due to the initial impact of the fault. The amount of recovery is determined by the recovery factor. control; Calculate the recovery amount: , where 0≤ ≤1, =1 means that the fault is completely reversible (such as a temporary problem solved by software restart). =0 means that the damage is permanent (such as plastic deformation of mechanical parts); Update health: After recovery, health must not exceed the level before the failure occurred. This aligns with the actual engineering requirements; Renewal of Vulnerability: Over time, the stress in a system slowly dissipates through material self-healing (such as annealing) or maintenance. This process is controlled by the forgetting factor γ, specifically... ,in This represents the accumulated vulnerability value after damage. A constant close to 1 (e.g., 0.999) represents the vulnerability decay rate per unit time, and γΔt ensures that the forgetting effect is independent of the time step Δt. The relationship between the forgetting factor and the half-life can be derived using the following formula: For example, if the forgetting_factor is set to 0.999999 in the JSON configuration file, the model will be updated once per second, and its implicit half-life will be calculated in reverse. This is approximately equivalent to 8 days, indicating that the model is configured with a memory length of about one week, and the impact of short-term disturbances will significantly decay within about a week. This calculation verifies whether the chosen parameters align with expectations for the system's memory duration. This design, which decouples recovery from forgetting, gives the model stronger explanatory power. It explains why a device that has undergone multiple major overhauls, even if its current HI score is the same as a new device, still has a higher risk of future failure because its vulnerability V(t) is much higher.

[0031] Step S301: Remaining life prediction At the end of each time step, a prediction is made based on the latest health status. Linear extrapolation method is used: ,in The failure threshold is used as the basis for this prediction. It is based on the time required for the current health level to drop to the failure threshold, intuitively reflecting the expected remaining lifespan of the equipment under the current health condition and natural wear rate. Since the impact of failure events is reflected in HI in real time, this prediction naturally incorporates the cumulative effect of sudden failures. When operating conditions need to be considered: , Where OC(t) represents operating condition parameters, such as load, temperature, and speed, which affect the rate of equipment degradation, and f( ) can be a linear model, neural network, Gaussian process regression, or other modeling methods, and e(t) is the modeling error or noise term, reflecting the uncertainty of the model.

[0032] This invention uses two state variables, health and vulnerability, to dynamically simulate the natural aging of equipment and the impact of random failures. Failure damage is amplified by the current vulnerability and accumulated in the vulnerability to form damage memory. After the failure is recovered, the health partially recovers, while the vulnerability decays exponentially through a forgetting factor. Finally, the remaining service life is predicted based on the current health and failure threshold, providing a reliable basis for predictive maintenance that combines dynamic response capability and long-term trend judgment.

[0033] Simulation verification: To visually demonstrate the effectiveness of this method, a circuit breaker was simulated for 100 days, with an initial HI=100 and V=0. Figure 1 The data is presented smoothly, showing the evolution of the device's health index (HI) and vulnerability level (V) over time. The horizontal axis represents time, the left vertical axis reflects the changes in HI, and the right vertical axis shows the trend of V. Figure 2 The diagram shows how the Remaining Lifetime (RUL) evolves over time, with the horizontal axis marking the number of days elapsed and the vertical axis reflecting the trend of RUL. When the equipment is running normally, as time Δt increases, the HI value decreases linearly according to the natural wear coefficient, while the fragility decreases exponentially with the forgetting factor. During this period, the machine operates very stably and has not experienced any failures for 9 consecutive days. The HI index gradually decreases and eventually achieves the preset target value. On day 10, the simulation event table showed an undervoltage alarm, and the system determined the severity level of the event to be S. According to the health damage calculation formula, this incident caused an immediate decrease in HI, and vulnerability accumulated accordingly, rising to the damage value level. On day 11, the fault was eliminated, and the system returned to normal operation, based on the duration effect function T. eff Combined with the fault recovery coefficient βr, HI recovers partially, but the recovered HI does not exceed the value before the fault occurred. Although the vulnerability is gradually forgotten over time, a cumulative effect still remains.

[0034] On day 31, an over-temperature alarm occurred again. After about 7 hours, the HI was deducted again, and the vulnerability increased. During the recovery process, the health partially recovered, but it still continued to decline.

[0035] On day 52, an overcurrent alarm occurred and was resolved after 4 hours. After the health status was restored, multiple failures had a cumulative effect. The curve shows that each failure event caused HI to drop sharply. Although there was a slight increase after recovery, the overall trend showed a sawtooth decline.

[0036] Finally, after the equipment had been running continuously for 100 days, the HI score dropped from the initial 100 to 55.51. Based on the preset failure threshold (20 points), the remaining service life (RUL) of the device was calculated to be 1284.46 days using linear extrapolation and failure accumulation model.

[0037] Example 2: This invention provides a system for assessing equipment health and predicting remaining life for implementing the method, the system comprising: Parameter configuration module: used to store and manage all model parameters, including natural attrition rate, forgetting factor, failure threshold, and severity, recovery factor, and duration effect function parameters of various faults. It loads the JSON configuration file as described in Example 1 through the interface. Status maintenance module: Connected to the parameter configuration module, it acts as the core memory and is responsible for maintaining and updating the device's health (HI) and vulnerability (V) in real time. Within each time step, it automatically performs the following: reads the natural wear rate from the parameter configuration unit and calculates the natural wear of HI; receives the damage or recovery amount from the event processing engine and updates the health and vulnerability (V); reads the forgetting factor and performs vulnerability decay. Event processing engine: Connected to the parameter configuration module and status maintenance module, it continuously monitors real-time data of the equipment (such as voltage and current), and has built-in fault diagnosis logic to determine the occurrence and resolution of faults. When a fault is detected, it obtains the current V value from the status maintenance unit, obtains the severity and duration effect function parameters of the corresponding fault from the parameter configuration unit, calculates the damage amount, and sends the results to the status maintenance unit to trigger a status update. When the fault is resolved, it calculates the damage amount or recovery amount and drives the status maintenance module to complete the status update.

[0038] Lifetime prediction module: Connected to the status maintenance module and parameter configuration module, it obtains the latest health status from the status maintenance module, obtains the failure threshold and natural wear rate from the parameter configuration unit, calculates the remaining lifetime, and outputs the results to the display interface or the upper management system.

[0039] Simulation module: Connected to all the above modules, it can load simulation parameters into the parameter configuration module, inject preset fault event sequences into the event processing engine, drive system operation, and synchronously collect health, vulnerability, and remaining life time series data generated by the status maintenance module and the life prediction module, and finally generate simulation curves for analysis.

[0040] The system of this invention collects equipment data in real time and automatically calculates equipment health, vulnerability and remaining life through an embedded dual degradation mechanism. It can also combine the prediction results with simulation and deduction to realize online perception of equipment status and life warning. It can provide quantitative data basis for the formulation and optimization of maintenance strategies and improve the accuracy and foresight of operation and maintenance management.

[0041] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for assessing equipment health and predicting remaining life based on a dual degradation mechanism, characterized in that, include: Initialize the device's state variables, which include health HI, used to characterize the device's real-time functional state, and vulnerability V, used to characterize the device's accumulated damage memory. Within each discrete time step: The health status HI is updated based on a preset natural attrition rate; If a fault event occurs within the current time step, the damage amount D is calculated based on the severity and duration of the fault event and the current vulnerability V, and the health HI is updated based on the damage amount D. At the same time, the damage amount is added to the vulnerability V. If the fault is resolved, calculate the recovery amount R of health based on the preset recovery factor, and update the health HI based on the recovery amount R; The vulnerability V is updated by decay based on a preset forgetting factor; Based on the updated health status HI and the preset failure threshold, the remaining useful life RUL of the device is calculated.

2. The method for equipment health assessment and remaining life prediction based on a dual degradation mechanism according to claim 1, characterized in that, The damage amount D includes the instantaneous impact damage at the time of the failure event and the cumulative damage during the duration of the failure.

3. The method for equipment health assessment and remaining life prediction based on a dual degradation mechanism according to claim 2, characterized in that, The formula for calculating the instantaneous impact damage is as follows: ;in, This is instantaneous impact damage. V represents the severity of the current fault, and V represents the vulnerability at the time the fault occurs.

4. The method for equipment health assessment and remaining life prediction based on a dual degradation mechanism according to claim 2, characterized in that, The cumulative damage is calculated based on the fault duration using a logarithmic growth model or a linear capping model.

5. The method for equipment health assessment and remaining life prediction based on a dual degradation mechanism according to claim 1, characterized in that, The formula for calculating the recovery amount R is: ,in, The recovery factor corresponding to the current fault type. and These represent the instantaneous impact damage and the continuous cumulative damage caused by the fault, respectively.

6. The method for equipment health assessment and remaining life prediction based on a dual degradation mechanism according to claim 1, characterized in that, The decay and update process of the vulnerability V is an exponential decay process, specifically including: The damage accumulated from the fault events within the current time step is added to the vulnerability V; The vulnerability V, which is superimposed with the damage amount, is exponentially calculated with the preset forgetting factor γ, and the result is used as the final vulnerability value at the end of the current time step, where γ is a constant between 0 and 1.

7. The method for equipment health assessment and remaining life prediction based on a dual degradation mechanism according to claim 1, characterized in that, The health status HI is updated based on the natural attrition rate α, specifically as follows: .

8. The method according to claim 1, characterized in that, The remaining service life RUL The calculation formula is: ,in This is the failure threshold.

9. A system for predicting the remaining life of equipment that implements the method of any one of claims 1-8, characterized in that, include: The parameter configuration module is used to load and manage system parameters, including natural wear rate, forgetting factor, failure threshold and parameters of various fault types, and provides parameter query interface for other modules. The status maintenance module, at each time step, stores and updates the device's health HI and vulnerability V based on the natural wear and tear rate and forgetting factor. The event processing engine responds to fault occurrence and resolution events, triggers and executes fault damage update and fault recovery update logic, and drives the state maintenance module to perform corresponding health HI and vulnerability V update calculations. The lifespan prediction module calculates the remaining lifespan (RUL) based on the health status (HI) in the status maintenance module.

10. The equipment remaining life prediction system according to claim 9, characterized in that, The system also includes a simulation module, which drives the system to run according to a preset sequence of fault events and parameters, and outputs simulation data and curves showing the changes in health (HI), vulnerability (V), and remaining lifetime (RUL) over time.

Images