A Bridge Component Inspection Strategy Planning Method Considering Spatial Variability
By using a dynamic Bayesian network model and spatially correlated nodes, the problem of unconsidered spatial variability in bridge component inspection was solved, enabling efficient and economical inspection strategy planning and ensuring component safety performance.
Patent Information
- Application Number
- CN202310549356.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2023-05-16
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2043-05-16
AI Technical Summary
Existing bridge component inspection strategies fail to effectively consider the spatial variability of components, resulting in inaccurate performance evaluation results, misleading decision-making, and failing to allocate maintenance resources reasonably, thus affecting the efficiency of inspections.
A dynamic Bayesian network model is adopted, spatial correlation nodes are introduced, the overall performance status is inferred through the local detection information of components, the inspection strategy threshold is set, the inspection frequency and location are reasonably adjusted, and the strategy is flexibly adjusted in combination with the detection information.
It enables efficient inspection based on the actual performance evolution trend of components, rationally plans inspection time intervals and locations, reduces unnecessary work, and improves the economy and safety of operation and maintenance management.
Smart Images

Figure CN116523254B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of bridge component inspection technology, and specifically to a bridge component inspection strategy planning method that takes into account spatial variability. Background Technology
[0002] To ensure the safety performance of bridge structures during their service life, maintenance departments conduct regular inspections and tests to promptly identify and repair defects in bridge components, preventing collapses. However, limited maintenance resources make the inherent costs of inspection activities a significant operational burden for maintenance departments. Therefore, rationally planning bridge component inspection strategies (inspection frequency, inspection locations, and testing technologies) is a key issue in the daily operation and maintenance management of bridge maintenance departments. Current engineering practices employ fixed-interval inspection strategies for bridge components, selecting a prescribed number of locations for testing. This approach fails to provide an overall understanding of the component's performance status and does not consider the maintenance needs at different stages of the component's lifespan, thus failing to rationally allocate maintenance resources throughout the bridge's life cycle.
[0003] Currently, research on bridge component inspection strategies both domestically and internationally mainly includes reliability-based methods, risk analysis methods, objective optimization methods, and uncertainty-based methods. While all of these methods involve simulating the degradation of component performance, they neglect the spatial variability of parameters within the model. Consequently, the performance evaluation results obtained may not be accurate, often leading to overestimation of reliability and misleading decision-makers, causing them to miss the optimal inspection time and limiting the efficiency of bridge inspection and maintenance to some extent. Therefore, there is an urgent need for an inspection strategy planning method that can consider the spatial variability of components. Summary of the Invention
[0004] The purpose of this invention is to propose a bridge component inspection strategy planning method that considers spatial variability. It introduces spatial correlation nodes based on dynamic Bayesian networks to consider the variability characteristics of random variables in the component degradation model. On the one hand, it can infer other location information from the local location detection information of the component, and then calculate the overall performance status of the component. On the other hand, by setting a performance evaluation threshold, it can reasonably consider the performance status of the component at different life stages, so as to adjust the component inspection frequency.
[0005] To achieve the above objectives, this application proposes a bridge component inspection strategy planning method that considers spatial variability, comprising:
[0006] Part 1: Selecting a component degradation model: Based on the bridge component defects of interest, select the corresponding degradation mechanism model;
[0007] Part Two: Establishing the Dynamic Bayesian Network Representation of the Component Degradation Model: The random variables in the component degradation model are classified and summarized into time-invariant node set I, time-varying node set C, and component performance state node set D. The prior probability distribution, conditional probability distribution, and probability distribution of the transition network of each node are determined. The static Bayesian network at the initial moment is established through the causal relationship between nodes, and then the transition network is used to extend it into a dynamic Bayesian network with the concept of life cycle.
[0008] Part Three, Introducing Spatial Relevance Nodes: The component is divided into n units using a unit discretization method, and each unit is represented by the dynamic Bayesian network.
[0009] Specifically, Gaussian-type functions are used to simulate the spatial correlation coefficients between different units:
[0010]
[0011] Where, ρ ij x is the correlation coefficient between nodes i and j that have spatial variation characteristics; i ,x j is the distance (mm) from the center point of element i and element j to the origin; b is the relevant length (mm);
[0012] Establish the correlation matrix R = [ρ] of the component elements. ij The dynamic Bayesian networks of different units are connected to form a basic model for evaluating the overall performance of the components.
[0013] To reduce computational burden, the vector composed of nodes with spatially variable characteristics in the dynamic Bayesian network is decomposed into a product of a transition matrix and a random vector:
[0014]
[0015] in, Represents the approximate vector after vector decomposition of the spatially variable node set (I or C); Let be the transformation matrix, and have S is a diagonal matrix introduced to reduce decomposition error; vectors U and V are independent normally distributed random variables, and have the following properties: (i = 1, 2, ..., m; j = 1, 2, ..., n, and m ≤ n), μ U μ V These are the sub-variables U within vectors U and V, respectively. i V i The mean; σ U σ V These are the sub-variables U within vectors U and V, respectively. i Vi The variance; m is the number of sub-variables that need to be retained in vector U during the decomposition process, and n is the number of units in the component division; the initial time μ U =μ V =0,σ U =σ V =1; In addition, vectors U and V are time-invariant vectors without being updated;
[0016] Therefore, node The mean is
[0017] In the decomposed dynamic Bayesian network, a binary discrete random variable node g (with values of 0 and 1) is introduced as a child node of the component unit performance status node D, to intuitively determine whether a component unit has failed:
[0018]
[0019] in, This indicates whether the i-th element of the component fails at time t. This indicates that the unit is not faulty. This indicates that the unit has failed; D represents the performance state value of the i-th element of the component at time t; cr This represents the limit value of a unit's performance state; exceeding this value indicates unit failure. Simultaneously, to enable the dynamic Bayesian network to be updated using detection information, the observation nodes... As child nodes;
[0020] To improve computing performance, nodes are introduced. Determine the performance status of different units at the nodes The connection simulates the entire component system as a series system:
[0021]
[0022] The binary discrete random variable node S is used to determine the overall performance of the component. t (Values are 0 and 1) as the node of the nth element of the component. The child nodes, and have Therefore, we obtain Figure 4 The bridge component degradation model based on dynamic Bayesian networks, considering spatial variability, is shown.
[0023] Part Four: Setting Inspection Strategy: Setting the maximum allowable inspection time interval Δt max (Year), Failure Probability Threshold P T Number of detection units per test (n) ins This is used to obtain inspection strategies for different types of bridge components;
[0024] Part 5: Updating the parameters of the bridge component degradation model based on inspection results: Using the bridge component inspection strategy and the bridge component degradation model, the failure probability of the bridge component within its lifespan under different strategies is obtained. When the failure probability exceeds a set threshold P... T Regularly inspect components;
[0025] The inspection results obtained during the inspection activities (whether the inspected unit is faulty or not) serve as nodes. child nodes The state is used to update the parameter distribution of each variable in the bridge component degradation model, as follows:
[0026] A1: Update nodes using Bayes' theorem probability distribution:
[0027]
[0028] in, Represents a node Relative to node The conditional probability; For nodes Relative to node The conditional probability; Represents a node At time-invariant nodes With time-varying nodes Prior probabilities under given conditions; Represents a node The prior probability;
[0029] A2: Invariant nodes during updates and time-varying nodes Parameter distribution:
[0030]
[0031] in, Represents the time-invariant node of the i-th unit at time t. or time-varying nodes p zt,i Represents variable nodes In the distribution of This conditionally equivalent quantile;
[0032]
[0033] in, For nodes Updated distribution mean; For nodes Since the variance of the random variable within a unit is small when the number of units is large enough, it is assumed that the variance will not change and no update is needed.
[0034] A3: Update the nodes representing spatial correlation. and Parameter distribution:
[0035]
[0036] Where k,l represents the subscript of the unit corresponding to any two inspection results among multiple detection unit results; s lj ,s kj t represents the element in the l-th row and j-th column of the error matrix S, and the element in the k-th row and j-th column, respectively; lj ,t kj Transformation matrices The element in the l-th row and j-th column, and the element in the k-th row and j-th column; thus, for the node and The mean of the distribution was estimated and updated once;
[0037] A4: Estimate based on q (2≤q≤n) inspection results obtained during the inspection activities. and
[0038]
[0039] in, For each node and The estimated value after multiple updates to the distribution mean; These are the nodes used for updating in step A3. and A function of the distribution mean;
[0040] Part VI. Predicting Component Performance: After the parameters of the bridge component degradation model at time t are updated, the component performance status at subsequent times is obtained;
[0041] First, from the updated nodes and Calculate the distribution values of the spatially variable nodes at the next time step (t+1):
[0042]
[0043] Secondly, calculate the probability that each element of the component has not failed:
[0044]
[0045] Furthermore, computing nodes The probability of:
[0046]
[0047] Finally, the overall failure probability of the component is calculated:
[0048]
[0049] The calculation of the overall failure probability of the component at subsequent times can be performed by repeating the steps in Part 6; at the same time, for times when the failure probability also exceeds the threshold, the calculation steps in Parts 5 and 6 can be repeated.
[0050] Part VII. Analysis of Inspection Strategy Benefits: After obtaining the performance prediction results within the structural lifecycle, the benefit values of different strategies are obtained, which are used for strategy comparison and analysis to obtain the optimal strategy; the benefit values considered in this invention include the inspection cost-benefit value and the failure risk benefit value; wherein, the inspection cost-benefit value is:
[0051]
[0052] in, Let be the total inspection cost-effectiveness value of the component from the initial time to time t; k is the total number of inspection activities performed before time t; t i This refers to the time point when the i-th inspection activity was previously performed. The immediate, undiscounted cost of implementing the i-th inspection activity; γ is the discount rate;
[0053] Secondly, the risk-benefit analysis for failure is as follows:
[0054]
[0055] Among them, U f C represents the total failure risk-benefit value of the component from the initial time to time t; f P(S) represents the immediate, undiscounted failure risk cost of the component in year i; t =0|S t-1 =1) represents the probability that the component will fail at time t assuming it has not failed at time (t-1);
[0056] Therefore, the total strategic benefit at any point in the component's lifecycle is:
[0057]
[0058] Part 8. Strategy Comparison and Analysis: After obtaining the strategy benefit values, the strategy with the lowest benefit value is selected as the optimal strategy through comparison.
[0059]
[0060] Where, π *π represents the optimal strategy; π represents all feasible strategies set.
[0061] Compared with existing technologies, the technical solution adopted in this invention has the following advantages: This invention introduces spatially correlated nodes into the bridge component degradation model based on a dynamic Bayesian network to consider the spatial variability of random variable parameters within the component degradation model. Considering spatial variability allows the component to infer variable information at its related locations from local detection information, thereby perceiving the overall performance status of the component. Simultaneously, by setting the maximum allowable inspection time interval, failure probability threshold, and number of detection units per instance, different component inspection strategies can be obtained. After comparing the total benefit values of the strategies, the optimal strategy can be selected, thus rationally planning the inspection time interval and the number of detection locations, avoiding unnecessary inspection work, and making operation and maintenance management economical and efficient. Furthermore, the inherent model update advantage of the dynamic Bayesian network used in this invention allows the method to flexibly adjust the inspection strategy based on detection information, which is beneficial for considering the actual performance evolution trend of the component. Attached Figure Description
[0062] Figure 1 This is a flowchart of a bridge component inspection strategy planning method that takes into account spatial variability;
[0063] Figure 2 This is a schematic diagram of the reinforced concrete beam bridge used in the embodiment;
[0064] Figure 3 This is a schematic diagram of a dynamic Bayesian network for a component degradation model;
[0065] Figure 4 This is a schematic diagram of the bridge component degradation model after introducing spatial correlation nodes;
[0066] Figure 5 This is a schematic diagram illustrating the failure probability of components under different inspection strategies;
[0067] Figure 6 This is a graph showing the calculated benefits of different strategies. Specific implementation methods
[0068] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to limit the application; that is, the described embodiments are only a part of the embodiments of this application, and not all of them.
[0069] Example 1
[0070] like Figure 1As shown, this embodiment provides a bridge component inspection strategy planning method considering spatial variability, including eight parts: selecting a component degradation model, establishing a dynamic Bayesian network representation of the component degradation model, introducing spatially relevant nodes, setting an inspection strategy, updating the bridge component degradation model parameters based on inspection results, predicting component performance, analyzing the benefits of the inspection strategy, and conducting a strategy comparison analysis. Figure 2 Taking a reinforced concrete simply supported beam subjected to chloride ion corrosion as an example, this study observes the rust expansion and cracking status at the bottom of the beam throughout its life cycle and formulates an inspection strategy accordingly. The basic information of the beam is: span 6000mm, cross-sectional dimensions 250mm × 600mm, concrete strength grade C40, water-cement ratio 0.42, and three Φ18 steel bars at the bottom of the cross-section. The specific implementation steps are as follows:
[0071] Step 1: Select the component degradation model: that is, the concrete cracking model caused by steel reinforcement corrosion and cracking.
[0072] Specifically, the time for steel bars to rust is:
[0073]
[0074] Where c is the thickness of the concrete cover (mm); D c The apparent diffusion coefficient of chloride ions (m / s) 2 );C cr Critical chloride ion concentration (kg / m³) 3 );C S Chloride ion concentration on concrete surface (kg / m³) 3 C0 represents the initial chloride ion concentration inside the concrete (kg / m³). 3 ); erf(·) is the error function;
[0075] The corrosion rate of the reinforcing steel is:
[0076] λ(t) = 0.0116k i,δ i cor (t)
[0077] Where, k i,δ i is the influence coefficient of corrosion current density; cor (t) represents the steel corrosion current density (μA / cm). 2 ), and i cor (t)=i cor (1)×0.85(t-t0) -0.29 , where {i cor (1) = [37.8(1-W / C)] 1.64 ] / c} represents the corrosion current density (μA / cm) of the reinforcing steel at the moment of initial corrosion. 2W / C is the water-to-binder ratio;
[0078] The critical corrosion depth of steel reinforcement when concrete cracks is:
[0079]
[0080] Where, d s f is the nominal diameter (mm) of the steel bar when it is not corroded; cu,k The compressive strength of concrete is the rated value (MPa).
[0081] Step 2: Establish a dynamic Bayesian network representation of the component degradation model, such as... Figure 3 As shown:
[0082] The time-invariant type node set I includes t0, c, C cr C S C0, a time-varying type node set C includes i cor (t); The component state node set D includes λ(t); The statistical characteristics of the relevant variable nodes and other parameters involved are shown in the following table:
[0083] Table 1. Statistical characteristics of internal variables and related parameters in the steel corrosion model.
[0084]
[0085] Step 3: Introduce spatial correlation nodes;
[0086] Specifically, the simply supported beam is evenly divided into 10 elements along its length, and each element corresponds to... Figure 3 The dynamic Bayesian network shown is illustrated. The correlation coefficients between different units are determined by the Gaussian function shown in the following equation:
[0087]
[0088] Where, ρ ij x is the correlation coefficient between unit i and unit j; i ,x j is the distance (mm) from the center point of element i and element j to the origin; b is the relevant length (mm), in this example, we take twice the dividing element length b = 1200;
[0089] The thickness of the concrete cover, *c*, is treated as a random variable with spatial variability and decomposed, with *m* = *n* = 10. Simultaneously, component element state judgment nodes *g*, *Z*, observation node *O*, and overall component state judgment node *S* are introduced after the component state node *D*. This allows for the establishment of... Figure 4 The diagram shown is a schematic representation of a bridge component degradation model considering spatial variability (for a simplified graphical representation, the time intervals are not indicated in the diagram). and (Connections between nodes);
[0090] Step 4: Set up the inspection strategy;
[0091] Specifically, the inspection strategy for this simply supported beam includes the maximum permissible inspection time interval Δt. max (Year), Failure Probability Threshold P T Number of detection units per test (n) ins This example uses the parameter values shown in the table below to set the inspection strategy:
[0092] Table 2 Inspection Strategy Plan Setting Table
[0093]
[0094] Step 5: Update the parameters of the bridge component degradation model based on the inspection results;
[0095] Specifically, under each inspection strategy, inspections are performed when the maximum allowable time interval or failure probability threshold is reached. Assume that the inspection result for each activity is O=1, and based on this, the network parameters are updated according to steps A1 to A4 shown in Part 5. Let inspection strategy n... ins =3,P T =0.003,I max The process of updating the parameter c of the first unit at time t=2 will be used as an example:
[0096] Suppose the conditional probability relationship between node g and node O is as follows:
[0097] P(O=1|g=1)=0.998, P(O=0|g=1)=0.002
[0098] P(O=1|g=0)=0.005, P(O=0|g=0)=0.995
[0099] node The posterior probability distribution is:
[0100]
[0101] node With condition D < D cr Equivalent quantiles (obtained through numerical solution) The probability distribution of ) is updated as follows:
[0102] parameter The mean is updated as follows:
[0103]
[0104] The other two inspection results are... The update steps are the same, and the results are 68.14 and 79.60 respectively;
[0105] The node is updated by three inspection results. and Distribution:
[0106]
[0107] Step Six: Predict Component Performance;
[0108] Specifically, based on the update results of spatially correlated nodes at the detection time, the failure probability of components at subsequent time points is calculated. Taking the update results at time t=2 in step five as an example, the intrinsic parameter values of the model at time t=3 are calculated:
[0109] First, parameters The mean calculation result is:
[0110]
[0111] Calculate the probability of each unit not failing:
[0112]
[0113] compute nodes The probability of:
[0114]
[0115] Calculate the overall failure probability of the component at this moment:
[0116]
[0117] Repeat step six to calculate the failure probability of the component at subsequent time points. When the failure probability threshold is exceeded, perform a routine inspection, and repeat steps five and six until the total failure probability over the component's lifecycle is obtained. The failure probability curves for different strategies in this example are shown below. Figure 5 As shown;
[0118] Step 7: Analyze the effectiveness of the inspection strategy;
[0119] Specifically, the benefit values of different strategies are calculated, including the cost-benefit value of inspection and the failure risk-benefit value. The cost-benefit value of inspection is:
[0120]
[0121] in, Let t be the total inspection cost-effectiveness value of the component from the initial time to time t; k is the total number of inspection activities performed before time t; t is the total number of inspection activities performed before time t. iThis refers to the time point when the i-th inspection activity was previously performed. The immediate, undiscounted cost of implementing the i-th inspection activity; γ is the discount rate;
[0122] The failure risk benefit is:
[0123]
[0124] Among them, U f C represents the total failure risk-benefit value of the component from the initial time to time t; f P(S) represents the immediate, undiscounted failure risk cost of the component in year i; t =0|S t-1 =1) represents the probability that the component will fail at time t assuming it has not failed at time (t-1);
[0125] Therefore, the total benefit of the strategy at any point in the component's lifecycle is:
[0126]
[0127] This example sets the detection cost of a single unit in a single inspection to a unit cost of 1, and the failure cost of a unit to 500. The benefit values for different strategies are as follows: Figure 6 As shown, “Cf-i” and “CI-i” represent the inspection cost-benefit and failure risk benefit when the maximum allowable inspection interval is i years, respectively.
[0128] Step 8: Strategy Comparison and Analysis;
[0129] Specifically, the strategy with the lowest benefit value is selected as the optimal strategy through comparison: for Figure 6 A comparative analysis of the strategies shown indicates that the optimal strategy when the failure probability threshold is 0.003 is n. ins =3,Δt max =2; the optimal strategy when the failure probability threshold is 0.005 is n ins =3,Δt max =2;
[0130] Therefore, the method described in this invention provides a bridge component inspection strategy that takes into account spatial variability, which can reasonably plan the inspection time interval and the number of inspection points to ensure the safety performance of the components.
[0131] The foregoing description of specific exemplary embodiments of the invention is for illustrative and explanatory purposes. These descriptions are not intended to limit the invention to the precise forms disclosed, and it will be apparent that many changes and variations can be made in accordance with the foregoing teachings. The exemplary embodiments were chosen and described in order to explain the specific principles of the invention and its practical application, thereby enabling those skilled in the art to implement and utilize various different exemplary embodiments of the invention, as well as various different choices and variations. The scope of the invention is intended to be defined by the claims and their equivalents.
Claims
1. A bridge component inspection strategy planning method considering spatial variability, characterized in that, include: Select a component degradation model; To establish a dynamic Bayesian network representation of the component degradation model: classify the nodes within the component degradation model, establish a static Bayesian network according to causal relationships, and introduce time slices to extend it into a dynamic Bayesian network; Introducing spatial correlation nodes: Gaussian-type functions are used to simulate the spatial correlation coefficients between different units. in, For inter-unit With unit The correlation coefficient between nodes exhibiting spatial variability; For unit and unit The distance from the center point to the origin; For the relevant length; Establish the correlation matrix R of the component elements. The dynamic Bayesian networks of different units are connected to form a basic model for evaluating the overall performance of the components. The vector composed of nodes with spatially variable features in a dynamic Bayesian network is decomposed into a product of a transition matrix and a random vector: in, Represents a set of nodes with spatial variability or The approximate vector after decomposition of the vector; Let be the transformation matrix, and have ; It is a diagonal matrix; vector and Let be a normally distributed random variable whose internal subvariables are independent, and have the following properties: ( And there are ), , They are vectors , Sub-variables , The mean; , They are vectors , Sub-variables , The variance; For vectors in the decomposition process The number of sub-variables that need to be retained. The number of elements to divide the component; initial time ;vector and It is a time-invariant vector without being updated. Therefore, node The mean is ; The decomposed dynamic Bayesian network introduces binary discrete random variable nodes. As a component unit performance status node The child nodes are used to visually determine whether a component element has failed: in, express Time component number Is the unit faulty? This indicates that the unit is not faulty. This indicates that the unit has failed; express Time component number Performance status values of each unit; This represents the limit value of a unit's performance state; exceeding this value indicates unit failure. Simultaneously, to enable the dynamic Bayesian network to be updated using detection information, the observation nodes... As child nodes; Introducing nodes Determine the performance status of different units at the nodes The connection simulates the entire component system as a series system: The binary discrete random variable nodes used to determine the overall performance of the component As the node of the nth element of the component The child nodes, and have Thus, a bridge component degradation model considering spatial variability based on dynamic Bayesian networks is obtained. Configure inspection policy: Set the maximum allowed inspection interval. Failure probability threshold Number of detection units per test This is used to obtain inspection strategies for different types of bridge components; Update the parameters of the bridge component degradation model based on the inspection results: A1: Update nodes using Bayes' theorem. probability distribution: in, Represents a node Relative to node The conditional probability; For nodes Relative to node The conditional probability; Represents a node At time-invariant nodes With time-varying nodes Prior probabilities under given conditions; Represents a node The prior probability; A2: Invariant nodes during updates and time-varying nodes Parameter distribution: in, represent Time of the first Time-invariant nodes of each unit or time-varying nodes ; Represents variable nodes In the distribution of This conditionally equivalent quantile; in, For nodes Updated distribution mean; For nodes The variance; A3: Update the nodes representing spatial correlation. and Parameter distribution: in, This indicates the subscript of the unit corresponding to any two inspection results among multiple inspection unit results; These are the error matrices. Inner Line number Column, No. Line number Column elements; Transformation matrices Inner Line number Column, No. Line number Column elements; A4: Based on information obtained during the inspection activities Individual inspection results estimated and : in, For each node and The estimated value after multiple updates to the distribution mean; , These are the nodes used for updating in step A3. and A function of the distribution mean; Predicting component performance: First, based on the updated nodes... and Calculate the next time step Distribution values of nodes with spatial variability: Secondly, calculate the probability that each element of the component has not failed: Furthermore, computing nodes The probability of: Finally, the overall failure probability of the component is calculated: The calculation of the overall failure probability of the component at subsequent time points is repeated in this step; at the same time, for time points where the failure probability also exceeds the threshold, the parameter update step of the bridge component degradation model is repeated. It also includes strategy comparison analysis: after obtaining the strategy benefit values, the strategy with the lowest benefit value is selected as the best strategy through comparison. in, Indicates the optimal strategy; This represents all the feasible strategies set.
2. The bridge component inspection strategy planning method considering spatial variability according to claim 1, characterized in that, It also includes analyzing the effectiveness of inspection strategies: after obtaining the performance prediction results within the structure's life cycle, the effectiveness values of different strategies are obtained, and the strategies are compared and analyzed to obtain the best strategy.
3. The bridge component inspection strategy planning method considering spatial variability according to claim 2, characterized in that, The benefit value includes the inspection cost-benefit value and the failure risk benefit value; wherein, the inspection cost-benefit value is: in, For the component from the initial time to The overall cost-effectiveness of inspections at any given moment; for The total number of inspection activities carried out before the specified time; For the previously executed The timing of the next inspection activity; To implement the first The immediate, undiscounted cost of the second inspection activity; The discount rate; Secondly, the risk-benefit analysis for failure is as follows: in, For the component from the initial time to The total failure risk-benefit value at any given time; For component number The immediate, undiscounted, and ineffective risk cost of the year; Indicates the component in Assuming the time does not expire The probability of failure at any given moment; Therefore, the total strategic benefit at any point in the component's lifecycle is: 。
Citation Information
Patent Citations
Dynamic degradation reliability evaluation method based on life information fusion
CN107766628A
Bridge full-life maintenance strategy optimization method based on semi-Markov decision process
CN112632677A
