Digital-twin-based deductive evaluation method for health status of industrial equipment

Abstract

A deductive evaluation solution for the health status of industrial equipment. Equipment in a scenario is modeled by means of digital twin technology; physical equipment is communicatively linked with a digital model, and a deductive evaluation is performed by means of operating status data of a digital equipment model, so as to obtain a deduced expected service life; and according to a real equipment operating service life and the deduced expected service life, feedback correction is performed on a status evaluation matrix. Thus, the aim of giving an early warning for equipment having an abnormal operating status is achieved, thereby improving the reliability and stability of equipment, and ensuring the operation and security of a data center.

Description

Industrial equipment health status deduction and assessment method based on digital twin Technical Field

[0001] The present invention relates to the field of digital twins, and in particular to a state deduction and evaluation method in digital twins. Background Art

[0002] Digital twin refers to the use of digital technology to build a "digital twin" model corresponding to the actual physical system in a virtual environment, and complete the simulation, analysis and prediction of the real world by interconnecting data with the actual physical system, thereby improving operational efficiency and accuracy in the real world.

[0003] In data center infrastructure scenarios (such as Industrial Internet applications), equipment failures can significantly impact the normal operation and security of data centers. Therefore, developing digital twin simulation and evaluation technology is crucial for ensuring the reliability and stability of data center equipment and improving operational efficiency and security.

[0004] Traditional monitoring methods often only issue alerts after equipment failures occur, failing to provide early warnings or detect abnormal changes in equipment health. Digital twin simulation and evaluation technology connects physical equipment to digital models, creating a virtual image of the physical equipment. Simulation and evaluation are performed using the operating status data of the digital equipment model, enabling early warning and fault analysis of equipment with abnormal operating conditions. Digital twin technology can also help optimize equipment design and operating parameters, improving performance and efficiency, thereby reducing energy consumption and maintenance costs in data centers.

[0005] In short, digital twin simulation and evaluation technology has broad application prospects in data center equipment fault warning and optimization management. It can improve the reliability, stability and operating efficiency of data centers, and provide important guarantees for the safe and stable operation of data centers.

[0006] Summary of the Invention

[0007] To address the above problems, the present invention provides an industrial equipment health status deduction and evaluation method based on digital twins, applying digital twins to data center scenarios (such as industrial Internet application scenarios) to achieve the purpose of improving the reliability, stability and operating efficiency of data centers, thereby providing a basis for the safe and stable operation of data centers.

[0008] The specific steps of the industrial equipment health status deduction and assessment method based on digital twins are as follows:

[0009] Step 1: Integrate digital twin technology to build a digital virtual model for the physical equipment in the data center scenario and communicate and interact with the physical equipment to ensure that the real-time operating status information and real-time environmental information of the physical equipment in the data center scenario are synchronized with the digital virtual model;

[0010] Step 2: Using the historical operating status information and environmental information of the physical equipment in the data center scene as the empirical sample data, the state sequence (S1, S2, ..., S k ), which is used to describe the different states of the equipment during operation. The equipment life corresponding to each state is (d1, d2, ..., d k ), and construct the state transition probability matrix P k×k And the judgment matrix X 6×4 ;

[0011] Step 3: Establish a status assessment formula based on the device's operating time (L), current intensity (I), and voltage level (U) as parameters, and taking into account environmental factors such as temperature (T), humidity (H), and air dust mass (A), with the device's health status as the result.

[0012] Step 4: According to the state evaluation formula, calculate the evaluation state value Q0 of the current equipment, and match the closest state in the state sequence to obtain the corresponding expected life of the equipment

[0013] Step 5: Based on Q0, the state transition probability matrix is ​​used to obtain the transfer state sequence, and a deduction is performed to obtain the state evaluation matrix Q1 of the first deduction, and the corresponding state probability matrix P is obtained. r1 , for each evaluation state value in Q1, match the closest state in the state sequence and obtain the corresponding equipment life expectancy

[0014] Step 6: Repeat the deduction in step 5 until the expected life of the equipment is less than the safety value d s , get the state evaluation matrix Q of the numth deduction num , and the corresponding state probability matrix P rnum , for Q num Each evaluation state value in the state sequence matches the closest state and obtains the corresponding equipment life expectancy

[0015] Step 7: Calculate the expected life of the equipment during the entire simulation process until the end of the simulation to obtain the expected life

[0016] Step 8: Based on the actual operating life of the equipment and life expectancy Feedback correction is performed on the state assessment matrix.

[0017] Furthermore, the state transition probability matrix P in step 2 k×k And the judgment matrix X 6×4 The specific expression is:

[0018] Among them, P(S j |S i ) indicates that the state S i Transfer to state S j The probability of X mn It means that the mth evaluation factor is at the nth level. The evaluation factors are L, I, U, T, H, and A respectively. The evaluation levels include [none, low, medium, high], 1≤m≤6, 1≤n≤4.

[0019] Furthermore, the expression of the state evaluation formula in step 3 is:

[0020] Among them, (A1, A2, A3, A4, A5, A6) = (L, U, I, T, H, A), D is the distance between the equipment and the safety line, D h and D l are the upper and lower safety limits, μ m is the expected safety value of the mth evaluation factor, K1 is the amplification factor, w m is the weight of the mth evaluation factor, and the evaluation factors are L, I, U, T, H, and A respectively.

[0021] Furthermore, the weight w of the mth evaluation factor m The expression is:

[0022] Where m = 1, 2, ..., 6, n = 1, 2, ..., 4, X mn It means that the mth evaluation factor is at the nth level. The evaluation factors are L, I, U, T, H, and A respectively. The evaluation levels include [none, low, medium, high].

[0023] Furthermore, step 4 is specifically as follows:

[0024] Among them, r0 is the state number closest to Q0 matched in the state sequence.

[0025] Furthermore, in step 5, the state sequence P is transferred r0 =P` r0, where P` r0 is the state transition probability matrix P k×k In the r0th row, r0 is the state number closest to Q0 in the state sequence.

[0026] Furthermore, the state evaluation matrix Q1 and the corresponding state probability matrix P in the first deduction in step 5 are r1 The expression is:

[0027] Among them, τ(1) represents the random value of the oscillation function τ for the first time, and τ is a Gaussian distribution τ~N(μ,σ 2 ) oscillation function, μ, σ 2 are the mean and expected variance of the oscillation function respectively;

[0028] For each state evaluation value in Q1, match the closest state in the state sequence with the corresponding equipment life expectancy Specifically:

[0029] in, is the kth element in Q1, express The value of the kth element in d th is the threshold of the equipment's service life, is the transpose of l1.

[0030] Furthermore, the state evaluation matrix Q of the num-th deduction in step 6 is num And the corresponding state probability matrix P rnum , equipment life expectancy The details are as follows:

[0031] in, Q num kth num elements, for l n 0 um kth num elements, τ(num) represents the random value of the num-th oscillation function τ, The state evaluation matrix Q for the num-1th deduction num-1 The kth num-1 OK, is the state probability matrix P rnum-1 The kth num-1 OK, Indicates l num The transpose of d thIt is the threshold of the equipment's service life.

[0032] Furthermore, in step 6, the life expectancy is deduced The expression is:

[0033] NUM is the total number of simulations performed by the end of the simulation.

[0034] Furthermore, in step 7, according to the actual operating life of the equipment and life expectancy Update the device service life threshold d according to the following formula th , to realize the feedback correction of the state evaluation matrix:

[0035] in, is the actual operating life of the equipment, new_d th This is the threshold value of the device's service life after the update. Beneficial effects

[0036] Traditional monitoring methods often only issue alarms after equipment failures occur, but are unable to provide early warnings or detect abnormal changes in equipment health. The digital twin deduction and evaluation method proposed in this patent connects physical equipment with digital models, constructs the digital model as a virtual image of the physical equipment, and performs deduction and evaluation based on the operating status data of the digital equipment model, thereby achieving early warning and fault analysis for equipment with abnormal operating status. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] FIG1 is a flow chart of the present invention. DETAILED DESCRIPTION

[0038] The technical solution of the present invention will be further described below with reference to the accompanying drawings and specific embodiments:

[0039] A method for evaluating the health status of industrial equipment based on digital twins is shown in Figure 1. The specific steps are as follows:

[0040] Step S101, combining digital twin technology, establishes a digital virtual model for the physical equipment in the data center scenario (such as the industrial Internet application scenario), and communicates and interacts with the physical equipment to ensure that the real-time operating status information and real-time environmental information of the physical equipment in the data center scenario are synchronized with the digital virtual model.

[0041] First, the equipment in the data center scenario is modeled, and the physical characteristics and operating conditions of the model are consistent with the physical equipment; second, real-time data of the environment and equipment is collected through sensors, and the data is converted into digital format for data interaction with the digital virtual model; then, the mathematical model is integrated with the real-time data converted into digital format to establish a digital twin model.

[0042] Step S102: Using the statistical histogram method of the empirical sample data (historical equipment operation status information and environmental data), a complete state sequence (S1, S2, ..., S k ) is used to describe the different states of the equipment during operation. k Refers to the kth state during the operation of the equipment. A device has a gradual evolution process from a healthy state to an abnormal state prone to failure. This process is divided into k states, and each of these k states has a corresponding equipment life, forming an equipment life sequence (d1, d2, ..., d k ), and construct the state transition probability matrix P k×k To describe the state transition relationship of the device and the judgment matrix X 6×4 The judgment matrix considers three influencing factors inside the device (device operating time L, device current intensity I, and device voltage level U) and three influencing factors of the external environment (temperature T, humidity H, and air dust quality A), and divides each influencing factor into four levels according to the degree of influence: [none, low, medium, high].

[0043] Among them, P(S j |S i ) indicates that the state S i Transfer to state S j The probability of , and the sum of the probabilities of any row is 1, that is X mn It means that the mth evaluation factor is at the nth level. The evaluation factors corresponding to each row are L, I, U, T, H, and A respectively. The evaluation levels include [none, low, medium, high], 1≤m≤6, 1≤n≤4.

[0044] Step S103: By monitoring the influencing factors of the internal state of the digital virtual model in real time, a state evaluation formula is established with the equipment running time L, equipment current intensity I, and equipment voltage level U as parameters, and taking into account the environmental influencing factors of temperature T, humidity H, and air dust quality A, with the equipment health state as the result. According to the formula, the current equipment evaluation state value Q0 is obtained, and the closest state in the state sequence is matched to obtain the corresponding expected life of the equipment.

[0045] Among them, (A1, A2, A3, A4, A5, A6) = (L, U, I, T, H, A), D is the distance between the equipment and the safety line, D h and D l is the upper and lower safety limits; μ m is the expected safety value of the assessment factor, which is obtained from historical data by expert experience; K1 is the amplification factor, which is used to map Q0 to the state sequence.

[0046] The weight w of the mth evaluation factor m From the judgment matrix X 6×4 By entropy weight method, we can get:

[0047] Calculate the corresponding weight w based on the information entropy of the factors affecting the internal state of each device m :

[0048] According to the obtained Q0, find the closest state in the complete state sequence and get the device life corresponding to the state

[0049] Among them, Q0 is closest to the r0th state in the state sequence, is the expected life span of the corresponding state, P` r0 is the r0th row in the state transition probability matrix.

[0050] Step S104: Based on Q0, the device's transition state sequence P is obtained according to the state transition probability matrix. r0 , perform one deduction and obtain the first deduction equipment evaluation state value matrix Q1, the matrix size is 1×k, and the corresponding state probability matrix P r1 , the matrix size is 1×k, find the closest state for each state value in Q1 and get the corresponding expected life of the equipment

[0051] Among them, τ(1) represents the random value of the oscillation function τ for the first time, and τ is a Gaussian distribution τ~N(μ,σ 2 ) oscillation function, μ, σ 2 are the mean and expected variance of the oscillation function, respectively, and are obtained by the following formula:

[0052] Among them, according to each state evaluation value in Q1, the closest state and the expected life of the corresponding state are found in the complete state sequence, and the expected life matrix l1 of the same size as Q1 is created to calculate the expected life The specific steps are as follows:

[0053] in, express The value of the kth element in d th is the threshold of the equipment lifespan, which is (d1, d2,…, d k ) in the life expectancy matrix is ​​less than d th is taken as 0; is the transpose of l1, is the expected life span corresponding to Q1.

[0054] Step S105, iterate step S104 and continue the deduction until the expected life of the equipment is less than the safety value d s According to the following tree branch structure deduction formula, the device deduction state matrix Q of the numth deduction is obtained: num , the matrix size is 1×k num , and the corresponding state probability matrix P rnum , the matrix size is 1×k num , for Q num For each state value in the complete state sequence, find the closest state and get the corresponding expected life of the device

[0055] Among them, τ(num) represents the random value taken by the numth time of the oscillation function τ, is the state evaluation matrix Q num-1 The kth num-1 OK, is the state probability matrix P rnum-1 The kth num-1 OK, Indicates l num The transpose of .

[0056] Each simulation will result in an expected lifespan of the device. Taking probability into account, the more simulations are performed, the smaller the expected lifespan of the device will be. Therefore, it will definitely be less than a safe value. At this time, the final number of simulations is obtained. num refers to any simulation in the process until the termination of the simulation, and Num is the number of simulations at the termination.

[0057] Step S106: Count the total expected life of the equipment during the deduction process until the end of the deduction to obtain the deduced expected life

[0058] Where NUM is the total number of simulations performed at the end of the simulation.

[0059] Step S107: according to the actual operating life of the equipment and life expectancy Feedback correction update th :

[0060] in, is the actual operating life of the equipment, new_d th To feedback the updated device lifespan threshold, after the update is completed, use new_d th Update the corresponding value in the evaluation formula and participate in the calculation.

[0061] It should be pointed out that the description of the above embodiments is only used to help understand the method of this application and its core idea. For ordinary technicians in this technical field, without departing from the principles of this application, several improvements and modifications can be made to this application, and these improvements and modifications are also within the scope of protection of the claims of this application.

Claims

1. The method for evaluating the health status of industrial equipment based on digital twins is characterized by: The specific steps of this method are as follows: Step 1: Combine digital twin technology to build a digital virtual model for the physical equipment in the data center scene, and communicate and interact with the physical equipment to ensure that the real-time operating status information and real-time environmental information of the physical equipment in the data center scene are synchronized with the digital virtual model; Step 2: Using the historical operating status information and environmental information of the physical equipment in the data center scenario as empirical sample data, the state sequence (S1, S2, …, S k ), S k It refers to the kth state in the operation process of the equipment, and the corresponding equipment life sequence is (d1, d2, ..., d k ), and construct the state transition probability matrix P k×k And the judgment matrix X 6×4 ; Step 3: Establish a status evaluation formula with equipment running time L, equipment current intensity I, equipment voltage level U as parameters, and consider environmental influencing factors such as temperature T, humidity H, and air dust quality A, with equipment health status as the result; Step 4: According to the status evaluation formula, calculate the evaluation status value Q0 of the current equipment, and match the closest status in the status sequence to obtain the corresponding expected life of the equipment. Step 5: Based on Q0, the state transition probability matrix is ​​used to obtain the state sequence, and a deduction is performed to obtain the first deduced state evaluation matrix Q1, and the corresponding state probability matrix P is obtained. r1 , for each evaluation state value in Q1, match the closest state in the state sequence and obtain the corresponding equipment life expectancy Step 6: Repeat the iterative step 5 until the expected life of the equipment is less than the safety value d s , get the state evaluation matrix Q of the numth deduction num , and the corresponding state probability matrix P rnum , for Q num Each evaluation state value in the state sequence matches the closest state and obtains the corresponding equipment life expectancy Step 7: Calculate the expected life of the equipment during the entire simulation process until the end of the simulation to obtain the expected life of the simulation Step 8: Based on the actual operating life of the equipment and life expectancy Feedback correction is performed on the state assessment matrix.

2. The method for evaluating the health status of industrial equipment based on digital twins according to claim 1 is characterized in that: The state transition probability matrix P in step 2 k×k And the judgment matrix X 6×4 The specific expression is: Among them, P(S j |S i ) indicates that the state S i Transfer to state S j The probability of 1≤i,j≤k,X mn It means that the mth evaluation factor is at the nth level. The evaluation factors are L, I, U, T, H, and A. The evaluation levels include [none, low, medium, and high]. 1≤m≤6, 1≤n≤4.

3. The method for evaluating the health status of industrial equipment based on digital twins according to claim 1 is characterized in that: The expression of the state evaluation formula in step 3 is: Among them, (A1, A2, A3, A4, A5, A6) = (L, U, I, T, H, A), D is the distance between the equipment and the safety line, D h and D l are the upper and lower safety limits, μ m is the expected safety value of the mth evaluation factor, K1 is the magnification factor, w m is the weight of the mth evaluation factor, and the evaluation factors are L, I, U, T, H, and A respectively.

4. The method for evaluating the health status of industrial equipment based on digital twins according to claim 1 is characterized in that: The weight w of the mth evaluation factor m The expression is: Where m = 1, 2, ..., 6, n = 1, 2, ..., 4, X mn It means that the mth evaluation factor is at the nth level. The evaluation factors are L, I, U, T, H, and A respectively. The evaluation levels include [none, low, medium, and high].

5. The method for evaluating the health status of industrial equipment based on digital twins according to claim 1 is characterized in that: Step 4 is as follows: Among them, r0 is the state number closest to Q0 matched in the state sequence.

6. The method for evaluating the health status of industrial equipment based on digital twins according to claim 1 is characterized in that: The transition state sequence in step 5 in, is the state transition probability matrix P k×k In the r0th row of , r0 is the state number closest to Q0 matched in the state sequence.

7. The method for evaluating the health status of industrial equipment based on digital twins according to claim 1 is characterized in that: The state evaluation matrix Q1 and the corresponding state probability matrix P of the first deduction in step 5 r1 The expression is: Among them, τ(1) represents the random value of the oscillation function τ for the first time, and τ is a Gaussian distribution τ~N(μ,σ 2 ) oscillation function, μ, σ 2 are the mean and expected variance of the oscillator function respectively; For each state evaluation value in Q1, match the closest state in the state sequence and the corresponding equipment expected life Specifically: in, is the kth element in Q1, express The value of the kth element in , d th is the device lifespan threshold, is the transpose of l1.

8. The method for evaluating the health status of industrial equipment based on digital twins according to claim 1 is characterized in that: The state evaluation matrix Q of the numth deduction in step 6 num And the corresponding state probability matrix P rnum , equipment life expectancy The details are as follows: in, Q num k num elements, for k num elements, τ(num) represents the random value taken by the numth oscillation function τ, The state evaluation matrix Q for the num-1th deduction num-1 The kth num-1 OK, is the state probability matrix P rnum-1 The kth num-1 OK, Indicates l num The transpose of th It is the threshold of the equipment service life.

9. The method for evaluating the health status of industrial equipment based on digital twins according to claim 1 is characterized in that: Derivation of life expectancy in step 6 The expression is: Among them, NUM is the total number of simulations performed by the end of the simulation.

10. The method for evaluating the health status of industrial equipment based on digital twins according to claim 1 is characterized in that: In step 7, the actual operating life of the equipment and life expectancy Update the device lifespan threshold d according to the following formula th , to realize the feedback correction of the state evaluation matrix: in, is the actual operating life of the equipment, new_d th It is the threshold of the service life of the updated equipment.

Images