Water turbine blade coating binding force evaluation method and device

Abstract

The invention discloses a water turbine blade coating binding force evaluation method and device, and relates to the technical field of hydroelectric generation. The method comprises the following steps: synchronously acquiring multi-modal nondestructive testing data of a coating area of a water turbine blade to be evaluated; inputting the multi-modal nondestructive testing data into the coating binding force prediction model for binding force prediction and integration to obtain a coating binding force spatial distribution diagram; the coating binding force spatial distribution diagram is input into a high-fidelity digital twin body for coupling simulation and life analysis, and a coating remaining life distribution diagram is obtained; inputting the coating binding force spatial distribution diagram and the coating residual life distribution diagram into an intelligent maintenance decision model for decision generation to obtain a blade maintenance decision; and carrying out visual display on the coating binding force spatial distribution diagram and the coating residual life distribution diagram, and generating a blade coating binding force evaluation report. The problems of low evaluation precision, single information dimension and disjunction between evaluation and decision in the prior art are solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of hydroelectric power generation, and in particular to a method and device for evaluating the bonding force of a water turbine blade coating. BACKGROUND

[0002] A water turbine is the core energy conversion equipment of a hydropower station. The blades of the water turbine are long-term operated in high-speed and sand-containing water flow, and the working conditions are harsh, which are prone to cavitation, abrasion and fatigue damage. In order to protect the blades and improve their efficiency, an anti-abrasion and anti-cavitation protective coating is usually coated on the surface of the blades. The performance of the coating, especially the bonding force between the coating and the blade substrate, directly determines the protection effect and the safe operation life of the blades. Once the coating peels off, not only will the substrate be exposed to the harsh environment to accelerate damage, but also the peeled fragments may cause secondary damage to other flow parts of the water turbine, leading to serious accidents.

[0003] Therefore, it is of great importance to accurately and efficiently evaluate the bonding force of the water turbine blade coating and to develop a scientific maintenance strategy on this basis to ensure the safe, stable and economic operation of the hydropower station.

[0004] The existing coating bonding force evaluation methods mainly have the following shortcomings: 1) Low evaluation accuracy: such as grid method, tensile method, etc., which need to sample or conduct destructive tests on the blades, and cannot comprehensively evaluate the in-service blades, and the representative of the sampling results is limited.

[0005] 2) Low evaluation accuracy: such as ultrasonic wave, eddy current, etc., which can usually only detect macroscopic defects such as interface debonding, and are not sensitive to early micro-characteristics of the bonding force degradation (such as micro-crack initiation and interface performance degradation), and it is difficult to quantitatively evaluate the bonding force grade.

[0006] 3) Evaluation and decision-making are disconnected: the traditional evaluation results are mostly qualitative or discrete data points, which are difficult to intuitively reflect the spatial distribution state of the bonding force on the entire blade surface. At the same time, there is a lack of effective data-driven model connection between the evaluation results and the life prediction and maintenance decision-making, resulting in that the development of the maintenance plan mainly depends on the expert experience, which has subjectivity and lag, and cannot realize intelligent and refined management in the whole life cycle.

[0007] In summary, it is urgent to develop a comprehensive technical solution which can accurately, quantitatively and comprehensively evaluate the bonding force of the water turbine blade coating, and intelligently predict the remaining life and automatically generate the optimal maintenance decision. SUMMARY

[0008] The present application provides a method and device for evaluating the bonding force of a water turbine blade coating. The present application solves the problems of low evaluation accuracy, single information dimension and disconnection between evaluation and decision-making in the prior art.

[0009] In a first aspect, the embodiments of the present application provide a method for evaluating the coating adhesion of a water turbine blade, the method comprising: synchronously collecting multi-modal non-destructive testing data of a coating area of the water turbine blade to be evaluated, the multi-modal non-destructive testing data comprising macroscopic images, microscopic images, acoustic emission signals, and thermal imaging maps; inputting the multi-modal non-destructive testing data into a coating adhesion prediction model pre-trained based on a physical information multi-modal fusion algorithm to perform adhesion prediction and integration, and obtaining a coating adhesion spatial distribution map; inputting the coating adhesion spatial distribution map as an initial condition into a high-fidelity digital twin pre-constructed based on a digital twin technology to perform coupling simulation and life analysis, and obtaining a coating residual life distribution map; inputting the coating adhesion spatial distribution map and the coating residual life distribution map into an intelligent maintenance decision model pre-constructed based on a reinforcement learning algorithm to generate a decision, and obtaining a blade maintenance decision; using the high-fidelity digital twin to visually display the coating adhesion spatial distribution map and the coating residual life distribution map, and combining the blade maintenance decision to generate a blade coating adhesion evaluation report.

[0010] The technical solutions provided by the embodiments of the present application at least bring the following beneficial effects: By synchronously collecting multi-modal data such as macroscopic / microscopic images, acoustic emission, and thermal imaging, and using a physical information fusion model for prediction, the coating adhesion state can be comprehensively and deeply characterized from different physical dimensions, the limitations of single detection technology are overcome, and high-precision evaluation from qualitative to quantitative and from point to plane is realized; the high-precision coating adhesion spatial distribution map is used as an initial condition of a digital twin, the damage evolution process of the coating under real working conditions is deduced through high-fidelity coupling simulation, the dynamic and spatial prediction of the coating residual life is realized, and a scientific basis is provided for preventive maintenance; a reinforcement learning model is introduced, multiple factors such as blade health status, power station operation, logistics support, and economic benefits are considered, the optimal maintenance decision can be autonomously learned and generated, the blindness and subjectivity of traditional decision-making methods are avoided, and the maximization of operation and maintenance benefits is realized; a complete technical closed loop from data acquisition, state evaluation, and life prediction to intelligent decision-making and report generation is constructed, and intuitive visual display is performed through a digital twin, which greatly improves the transparency and efficiency of operation and maintenance work.

[0011] In an optional implementation, for a coating area of a water turbine blade to be evaluated, multi-modal non-destructive testing data is synchronously collected, comprising: based on a laser scanning technology, a three-dimensional point cloud model of the water turbine blade to be evaluated is obtained, and a global coordinate system of the water turbine blade is established; On the three-dimensional point cloud model, a scanning path covering the coating area of the whole blade surface is planned, and a plurality of detection points are set on the blade surface of the water turbine blade according to the scanning path; The multi-modal data acquisition probe is spatially registered in the global coordinate system to obtain the spatially registered multi-modal data acquisition probe; The multi-modal data acquisition probe is used to synchronously collect multi-modal data at each detection point to obtain multi-modal nondestructive testing data of the coating area of the water turbine blade to be evaluated.

[0012] In an optional implementation, the coating adhesion force prediction model comprises an image feature extraction branch and a time sequence feature extraction branch connected in parallel, and an attention fusion module, a physical parameter output layer and an adhesion force prediction layer connected in sequence; The image feature extraction branch comprises a macro image feature extractor constructed based on a ResNet-50 algorithm and a micro image feature extractor based on an EfficientNet-B0 algorithm connected in parallel, and a global average pooling layer; The time sequence feature extraction branch comprises an acoustic time sequence feature extractor constructed based on an MLP algorithm and a thermal imaging time sequence feature extractor based on a 3D-CNN algorithm connected in parallel; The coating adhesion force prediction model is provided with a total loss function combining physical information and data loss, the total loss function comprising a total physical loss function and a data loss function, the total physical loss function comprising a heat conduction physical loss function, a fatigue damage physical loss function and a physical consistency loss function, and the formula is:

[0013] In the formula, is the total loss function; is the data loss function; is the total physical loss function; , is the total loss weight;

[0014] In the formula, is the heat conduction physical loss function; is the fatigue damage physical loss function; is the physical consistency loss function; , , is the total physical loss weight;

[0015] In the formula, is the MSE function; The predicted surface temperature response curve is obtained by calculating the interfacial thermal resistance output by the output layer based on the physical parameters. This is the measured surface temperature response curve;

[0016] In the formula, This is a function for calculating the energy of acoustic emission signals. The energy of the acoustic emission signal; This is the mapping function from the stress factor to the damage space; The range of stress intensity factors;

[0017] In the formula, For interfacial thermal resistance; Microcrack density; Based on the thermal resistance of the interface; The damage-thermal resistance coupling coefficient; The intelligent maintenance decision model is built based on the PPO algorithm and includes an intelligent agent and an experience replay pool.

[0018] In one alternative implementation, the method for constructing the coating adhesion prediction model includes: Several coated test blocks of turbine blades simulating different working conditions were prepared in the laboratory. For each coated test block, multimodal nondestructive testing data were collected. For each coated specimen, destructive testing and surface temperature testing were performed to obtain the corresponding true value label of adhesion force and the measured surface temperature response curve, which were then added to the corresponding sample multimodal nondestructive testing data to obtain the sample dataset. An initial coating adhesion prediction model is constructed based on a physical information multimodal fusion algorithm. Based on the total loss function, a quantum-enhanced intelligent optimizer is used to optimize the hyperparameters of the initial coating adhesion prediction model, resulting in the final coating adhesion prediction model.

[0019] In one alternative implementation, based on the total loss function, a quantum-enhanced intelligent optimizer is used to optimize the hyperparameters of the initial coating adhesion prediction model, resulting in the final coating adhesion prediction model, including: The hyperparameter combination of the initial coating adhesion prediction model is associated with the QAOA parameters of the quantum-enhanced intelligent optimizer, and the total loss function is used as the cost function of the QAOA parameters. The quantum-enhanced intelligent optimizer is initialized using a Logistic chaotic map to generate initial QAOA parameters and then enters a hybrid iterative optimization loop. In each mixing iteration optimization cycle, on the quantum computer, the QAOA circuit is run using the current QAOA parameters to obtain the current quantum state, and the current hyperparameter combination is sampled; On the classical computer, using the sample data set, the current hyperparameter combination is used to quickly train the alternative coating adhesion force prediction model, and the cost function is used to calculate the current cost value of the alternative coating adhesion force prediction model; A flight random number is generated, if the flight random number is less than a preset threshold, Levy flight exploration is performed, the current QAOA parameters are updated to obtain updated QAOA parameters, otherwise, the gradient descent is used to update the current QAOA parameters to obtain updated QAOA parameters, and the hyperparameter combination obtaining and cost value calculating step is returned; If the current mixing iteration number reaches the mixing iteration number threshold or the cost value improvement is lower than the improvement threshold, the mixing iteration optimization step is terminated, and the optimal hyperparameter combination corresponding to the optimal cost value is output; According to the optimal hyperparameter combination, the hyperparameters of the initial coating adhesion force prediction model are set to obtain the final coating adhesion force prediction model.

[0020] In an optional implementation, the multi-modal non-destructive testing data is input into the coating adhesion force prediction model pre-trained based on the physical information multi-modal fusion algorithm for adhesion force prediction and integration to obtain a coating adhesion force spatial distribution map, including: The multi-modal non-destructive testing data is pre-processed to obtain pre-processed multi-modal non-destructive testing data, and is input into the coating adhesion force prediction model pre-trained based on the physical information multi-modal fusion algorithm; The macroscopic image feature extractor of the coating adhesion force prediction model is used to extract the macroscopic feature map of the macroscopic image in the pre-processed multi-modal non-destructive testing data, and input into the global average pooling layer for global average pooling to obtain the macroscopic image feature; The microscopic image feature extractor of the coating adhesion force prediction model is used to extract the microscopic feature map of the microscopic image in the pre-processed multi-modal non-destructive testing data, and input into the global average pooling layer for global average pooling to obtain the microscopic image feature; The acoustic time series feature extractor of the coating adhesion force prediction model is used to extract the acoustic time series feature of the acoustic emission signal in the pre-processed multi-modal non-destructive testing data; The thermal imaging time series feature extractor of the coating adhesion force prediction model is used to extract the thermal imaging time series feature of the thermal imaging map in the pre-processed multi-modal non-destructive testing data; The macroscopic image features, the microscopic image features, the acoustic time sequence features and the thermal imaging time sequence features are spliced to obtain a fusion feature vector, and the fusion feature vector is input into an attention fusion module of the coating adhesion prediction model to perform attention fusion to obtain a weighted fusion feature; According to the weighted fusion feature, a physical parameter output layer of the coating adhesion prediction model is used to generate a physical parameter; the physical parameter includes an interfacial thermal resistance, a microcrack density and an interfacial bonding strength; According to the physical parameter, a bonding force prediction layer of the coating adhesion prediction model is used to call a Softmax function to generate a bonding force grade, and the bonding force grade is combined with the corresponding physical parameter to obtain a bonding force prediction result; According to the spatial coordinate value of each detection point in the global coordinate system, the bonding force prediction results of all detection points are integrated to obtain a corresponding coating adhesion spatial distribution map.

[0021] In an optional implementation, the coating adhesion spatial distribution map is input into a high-fidelity digital twin body pre-constructed based on a digital twin technology to perform coupling simulation and life analysis, and a coating residual life distribution map is obtained, including: A three-dimensional point cloud model of the water turbine blade to be evaluated is imported into a simulation software, and a high-quality hexahedral dominant mesh is divided to obtain a corresponding three-dimensional simulation model; Based on the digital twin technology, an elastic-plastic material model of the matrix and the coating is defined in the three-dimensional simulation model, and a damage evolution equation corresponding to the elastic-plastic material model of the coating is set to obtain a pre-constructed high-fidelity digital twin body; The coating adhesion spatial distribution map is mapped to the high-fidelity digital twin body as an initial condition of the damage evolution equation, and the high-fidelity digital twin body is calibrated; The high-fidelity digital twin body is subjected to coupling simulation and damage accumulation under simulation of a specific working condition to obtain a damage value of each detection point; If the damage value exceeds a damage threshold for the first time, the detection point is taken as a key detection point, and a coupling simulation time experienced by the key detection point is taken as a corresponding predicted residual life; If the damage value does not exceed the damage threshold, the coupling simulation and damage accumulation are continued based on the detection point; According to the spatial coordinate value of each key detection point in the global coordinate system, the predicted residual lives of all key detection points are integrated to obtain a coating residual life distribution map.

[0022] In an optional implementation, the coating adhesion spatial distribution map and the coating residual life distribution map are input into an intelligent maintenance decision model pre-constructed based on a reinforcement learning algorithm to generate a decision, and a blade maintenance decision is obtained, including: extract global statistical features, local key region features and spatial context features of the coating bonding force spatial distribution map and the coating residual life distribution map, and combine the hydropower station state data to obtain complete state observation; write the complete state observation into the state space of the agent of the intelligent maintenance decision model pre-constructed based on the reinforcement learning algorithm; based on the state space, use the agent to select the optimal action in the preset action space to obtain the blade maintenance decision, and collect real-time experience in the decision generation process; extract a number of historical experiences from the experience replay pool, and combine the real-time experience to continuously train the intelligent maintenance decision model to obtain an updated intelligent maintenance decision model.

[0023] In an optional implementation, using a high-fidelity digital twin, the coating bonding force spatial distribution map and the coating residual life distribution map are visually displayed, and combined with the blade maintenance decision, a blade coating bonding force evaluation report is generated, including: load the high-fidelity digital twin of the water turbine blade to be evaluated; superimpose and render the coating bonding force spatial distribution map and the coating residual life distribution map as independent layers in the form of heat maps of different colors on the high-fidelity digital twin, and visually display them; write the coating bonding force spatial distribution map, the coating residual life distribution map and the corresponding blade maintenance decision into a preset report template to generate a blade coating bonding force evaluation report and visually display it.

[0024] In a second aspect, the embodiments of the present application provide a water turbine blade coating bonding force evaluation device for implementing the water turbine blade coating bonding force evaluation method. The device comprises: a multi-modal data acquisition unit for synchronously acquiring multi-modal non-destructive testing data of the coating area of the water turbine blade to be evaluated; a coating bonding force prediction unit for inputting the multi-modal non-destructive testing data into a coating bonding force prediction model pre-trained based on a physical information multi-modal fusion algorithm for bonding force prediction and integration to obtain a coating bonding force spatial distribution map; a coating residual life analysis unit for inputting the coating bonding force spatial distribution map as an initial condition into a high-fidelity digital twin pre-constructed based on digital twin technology for coupling simulation and life analysis to obtain a coating residual life distribution map; a blade maintenance decision generation unit for inputting the coating bonding force spatial distribution map and the coating residual life distribution map into an intelligent maintenance decision model pre-constructed based on a reinforcement learning algorithm for decision generation to obtain a blade maintenance decision; The visualization display unit is used for visualizing the coating adhesion force spatial distribution map and the coating residual life distribution map using a high-fidelity digital twin, and generating a blade coating adhesion force evaluation report in combination with a blade maintenance decision.

[0025] The third aspect of the embodiment of the present application provides an electronic device, and the electronic device comprises: at least one processor; and a memory connected with the at least one processor in communication; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor to enable the at least one processor to perform the method according to the first aspect of the embodiment of the present application.

[0026] The fourth aspect of the embodiment of the present application provides a computer readable storage medium, and the computer readable storage medium stores a computer program, and the computer program is executed by a processor to implement the method according to the first aspect of the embodiment of the present application. BRIEF DESCRIPTION OF DRAWINGS

[0027] Figure 1 The electronic device structure schematic diagram of the hardware running environment related to the embodiment of the present application is shown in the figure. Figure 2 The step flow chart of the water turbine blade coating adhesion force evaluation method provided by the embodiment of the present application is shown in the figure. Figure 3 The functional unit schematic diagram of the water turbine blade coating adhesion force evaluation device provided by the embodiment of the present application is shown in the figure. DETAILED DESCRIPTION

[0028] In order to make the above objectives, characteristics and advantages of the present application more apparent, comprehensible and easy to understand, the present application will be further described in detail below with reference to the drawings and specific embodiments. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0029] The scheme of the present application will be further described below with reference to the drawings.

[0030] Reference Figure 1 , Figure 1 The electronic device structure schematic diagram of the hardware running environment related to the embodiment of the present application is shown in the figure.

[0031] As Figure 1As shown, the electronic device can include: a processor 1001, for example, a Central Processing Unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, a memory 1005. Among them, the communication bus 1002 is used to realize the connection communication between these components. The user interface 1003 can include a display screen (Display), an input unit such as a keyboard (Keyboard), and the optional user interface 1003 can also include a standard wired interface, a wireless interface. The network interface 1004 can optionally include a standard wired interface, a wireless interface (such as a Wireless-Fidelity (WIreless-FIdelity, WI-FI) interface). The memory 1005 can be a high-speed Random Access Memory (RAM) memory, or a stable Non-Volatile Memory (NVM), such as a disk memory. The memory 1005 can also be an independent storage device from the aforementioned processor 1001.

[0032] Those skilled in the art can understand that Figure 1 The structure shown in the figure does not constitute a limitation on the electronic device, and can include more or fewer components than the figure, or combine certain components, or different component arrangements.

[0033] As Figure 1 As shown, the memory 1005 as a storage medium can include an operating system, a data storage module, a network communication module, a user interface module, and an electronic program of the water turbine blade coating bonding force evaluation device.

[0034] In Figure 1 As shown in the electronic device, the network interface 1004 is mainly used for data communication with the network server; the user interface 1003 is mainly used for data interaction with the user; the processor 1001 and the memory 1005 in the electronic device of the present application can be arranged in the electronic device, and the electronic device calls the electronic program of the water turbine blade coating bonding force evaluation device stored in the memory 1005 through the processor 1001, and executes the water turbine blade coating bonding force evaluation method provided by the embodiment of the present application.

[0035] Referring to Figure 2 , the embodiment of the present application provides a water turbine blade coating bonding force evaluation method, the method comprising: S201: synchronously collecting multi-modal non-destructive testing data on the coating area of the water turbine blade to be evaluated; the multi-modal non-destructive testing data includes macroscopic images, microscopic images, acoustic emission signals and thermal imaging spectra; S202: input the multi-modal non-destructive testing data into the coating adhesion prediction model pre-trained based on the physical information multi-modal fusion algorithm for adhesion prediction and integration, to obtain a coating adhesion spatial distribution map; S203: input the coating adhesion spatial distribution map as an initial condition into the high-fidelity digital twin pre-constructed based on the digital twin technology for coupling simulation and life analysis, to obtain a coating residual life distribution map; S204: input the coating adhesion spatial distribution map and the coating residual life distribution map into the intelligent maintenance decision model pre-constructed based on the reinforcement learning algorithm for decision generation, to obtain a blade maintenance decision; S205: use the high-fidelity digital twin to visually display the coating adhesion spatial distribution map and the coating residual life distribution map, and combine the blade maintenance decision to generate a blade coating adhesion evaluation report.

[0036] The technical scheme provided by the embodiments of the present application at least brings the following beneficial effects: By synchronously collecting multi-modal data such as macro / micro images, acoustic emission, thermal imaging, etc., and using a physical information fusion model for prediction, the coating adhesion state can be comprehensively and deeply characterized from different physical dimensions, overcoming the limitations of single detection technology and realizing high-precision evaluation from qualitative to quantitative and from point to area; the high-precision coating adhesion spatial distribution map is used as an initial condition of the digital twin, and the damage evolution process of the coating under real working conditions is deduced through high-fidelity coupling simulation, realizing dynamic and spatial prediction of the coating residual life and providing a scientific basis for preventive maintenance; the reinforcement learning model is introduced, and multiple factors such as blade health status, power station operation, logistics support and economic benefits are considered, which can autonomously learn and generate the optimal maintenance decision, avoiding the blindness and subjectivity of traditional decision-making methods and maximizing the operation and maintenance benefits; a complete technical closed loop from data acquisition, state evaluation, life prediction to intelligent decision-making and report generation is constructed, and intuitive visual display is realized through the digital twin, greatly improving the transparency and efficiency of operation and maintenance work.

[0037] In an optional implementation, multi-modal non-destructive testing data of the coating area of the water turbine blade to be evaluated are synchronously collected, including: S2011: based on laser scanning technology, a three-dimensional point cloud model of the water turbine blade to be evaluated is obtained, and a global coordinate system of the water turbine blade is established; In the embodiment, the detailed steps include: A ground-based or handheld three-dimensional laser scanner is used to perform omnidirectional scanning on the water turbine blade in a stopped state, to obtain original point cloud data containing millions of data points; Use point cloud processing software (such as Geomagic, PolyWorks) to denoise, register (align the data of multiple scans), and encapsulate the original point cloud data, and generate a continuous, flawless three-dimensional point cloud model (STL or OBJ format) of the water turbine blade to be evaluated; Define a global three-dimensional Cartesian coordinate system (O-XYZ) on the three-dimensional point cloud model, that is, the global coordinate system; usually, the origin O can be set at the center of the root flange of the water turbine blade to be evaluated, the X axis can point to the outer edge of the blade along the radial direction, and the Z axis can point along the main shaft direction of the water turbine; all subsequent spatial data will be aligned with this coordinate system; S2012: On the three-dimensional point cloud model, plan a scanning path covering the entire coating area of the blade surface, and set a number of detection points on the blade surface of the water turbine blade according to the scanning path; In this embodiment, the detailed steps include: Run the path planning algorithm on the three-dimensional point cloud model; the algorithm automatically generates one or more "S" or spiral scanning paths that can efficiently cover the entire surface based on the curvature distribution of the blade and the stress distribution history (if any); In areas with high curvature and stress concentration (such as the water inlet edge and the water outlet edge), the path density is higher and the detection point spacing is smaller (such as 5 cm); in flat areas, the spacing can be appropriately increased (such as 10 cm); Generate a series of discrete detection points along the planned path at fixed intervals, and each detection point has its precise coordinates and surface normal vector in the global coordinate system; S2013: Spatially register the multi-modal data acquisition probe in the global coordinate system to obtain the spatially registered multi-modal data acquisition probe; In this embodiment, the detailed steps include: Integrate an industrial camera, a long working distance microscope, an acoustic emission sensor array, an infrared thermal imager, an electromagnetic exciter, etc. in a compact multi-modal data acquisition probe shell; a high-precision positioning target (such as a reflective marker ball) is installed at the end of the multi-modal data acquisition probe; Through the "hand-eye calibration" process, the transformation relationship between the coordinate systems of the sensors inside the probe and the coordinate system of the positioning target is accurately determined; Install the multi-modal data acquisition probe on a robot arm or a rail system; install one or more tracking cameras (such as a laser tracker) beside the water turbine blade to be evaluated; the tracking camera tracks the positioning target on the probe in real time, thereby calculating the precise pose (position and attitude) of the end of the probe (i.e. the center of the sensor) in the global coordinate system O-XYZ in real time; this enables each frame of data collected to be accurately assigned a three-dimensional spatial coordinate; S2014: Using a multimodal data acquisition probe, a standardized multimodal data acquisition process is executed sequentially at each detection point to synchronously acquire multimodal data and obtain multimodal nondestructive testing data of the coating area of ​​the turbine blade to be evaluated; In this embodiment, the detailed steps include: The robotic arm moves the multimodal data acquisition probe to the detection point and adjusts the probe's orientation according to the surface normal vector to make it perpendicular to the surface being measured; the central controller sends a synchronization trigger signal. Macroscopic and microscopic image acquisition: Macroscopic image: An industrial camera captures an image covering an area of ​​approximately 10cm x 10cm under uniform illumination from a ring-shaped LED light source, with a resolution of no less than 24 million pixels; Microscopic images: A high-magnification optical microscope (e.g., 200x) is aimed at the central area of ​​the macroscopic image to capture the microscopic morphology of the coating surface, such as pores, microcracks, and particle distribution; Acoustic emission signal acquisition: Excitation: The electromagnetic exciter generates a pulsed mechanical wave with a frequency of 20kHz and a duration of 5ms, which applies a small dynamic stress to the coating. Acquisition: Four acoustic emission sensors (acoustic emission sensor array, sampling rate set to 5MHz) arranged around the probe synchronously acquire acoustic emission signals within 50ms after excitation; the sound source can be initially located by using the time difference of arrival algorithm; Thermal imaging spectral acquisition: Excitation: Two high-energy pulsed xenon lamps (approximately 6kJ of energy) are aimed at the detection area and subjected to a 2ms flash heating. Acquisition: An infrared thermal imager (spectral range 3-5μm, thermal sensitivity <20mK, frame rate 200Hz) recorded the complete temperature change sequence from the start of heating to the end of cooling (approximately 10 seconds); The spatial coordinates, timestamp, and all collected raw data of the detection point are packaged to obtain multimodal nondestructive testing data, which is then stored in the database.

[0038] In one alternative implementation, the coating adhesion prediction model includes parallel image feature extraction branches and temporal feature extraction branches, as well as sequentially connected attention fusion module, physical parameter output layer and adhesion prediction layer; The image feature extraction branch includes a macroscopic image feature extractor constructed in parallel based on the 50-layer Residual Network 50 (ResNet-50) algorithm and a microscopic image feature extractor based on the EfficientNet-B0 algorithm, as well as a global average pooling layer. The temporal feature extraction branch includes a parallel acoustic temporal feature extractor based on the Multi-Layer Perceptron (MLP) algorithm and a thermal imaging temporal feature extractor based on the 3-Dimensional Convolutional Neural Network (3D-CNN) algorithm; The coating adhesion prediction model incorporates a total loss function that combines physical information and data loss. This total loss function includes a total physical loss function and a data loss function. The total physical loss function comprises a heat conduction physical loss function, a fatigue damage physical loss function, and a physical consistency loss function, as shown in the formula:

[0039] In the formula, This is the total loss function; For data loss function; This is the total physical loss function; , Weighted by the total loss;

[0040] In the formula, This is the physical loss function for heat conduction; The physical loss function is the fatigue damage function. The physical consistency loss function; , , This is the weight for the total physical loss; For thermal imaging data, a one-dimensional transient heat conduction model is established, with the following formula:

[0041] In the formula, For temperature; x Spatial location; t For time; pc It is the volumetric heat capacity; k Thermal conductivity; This represents the rate of change of temperature with respect to time. For temperature gradient; The divergence of heat flow; The symbol is for partial differentials; Among them, at the coating-substrate interface At this point, there exists an interfacial thermal resistance. The resulting temperature transition, among which, d For the interface position, the formula is:

[0042] In the formula, Temperature on the interface coating side; This refers to the temperature on the substrate side of the interface. This is a temperature transition; The heat flux density at the interface; The interfacial thermal resistance predicted by the coating adhesion prediction model As known parameters, the partial differential equation is solved using the finite difference method to obtain the predicted surface temperature response curve. ,calculate Response curve with measured surface temperature The mean square error between them is given by the formula:

[0043] In the formula, This is the Mean Squared Error (MSE) function; The predicted surface temperature response curve is obtained by calculating the interfacial thermal resistance output by the output layer based on the physical parameters. This is the measured surface temperature response curve; For acoustic emission signals, their energy It can be regarded as the increase in damage Relatedly, according to Paris's law in fracture mechanics, the damage increment is related to the range of stress intensity factors. Related; by using a simplified blade stress model or finite element simulation, the stress amplitude at the detection point is estimated, and then the range of stress intensity factor is calculated. ;

[0044] In the formula, This is a function for calculating the energy of acoustic emission signals. The energy of the acoustic emission signal; This is the mapping function from the stress factor to the damage space; The range of stress intensity factors;

[0045] In the formula, For interfacial thermal resistance; Microcrack density; Based on the thermal resistance of the interface; The damage-thermal resistance coupling coefficient; The intelligent maintenance decision model is built on the Proximal Policy Optimization (PPO) algorithm, and the intelligent maintenance decision model is equipped with an agent and an experience replay pool.

[0046] In one alternative implementation, the method for constructing the coating adhesion prediction model includes: A-1: In the laboratory, several coating test blocks of water turbine blades under different working conditions are prepared, and for each coating test block, sample multi-modal non-destructive testing data is collected; A-2: For each coating test block, destructive testing and surface temperature testing are performed to obtain the corresponding bond strength true value label and measured surface temperature response curve, and are added to the corresponding sample multi-modal non-destructive testing data to obtain a sample data set; A-3: Based on a physical information multi-modal fusion algorithm, an initial coating bond strength prediction model is constructed; A-4: Based on a total loss function, a quantum-enhanced intelligent optimizer is used to optimize the hyperparameters of the initial coating bond strength prediction model to obtain a final coating bond strength prediction model.

[0047] In an optional implementation, based on a total loss function, a quantum-enhanced intelligent optimizer is used to optimize the hyperparameters of the initial coating bond strength prediction model to obtain a final coating bond strength prediction model, including: A-41: The hyperparameter combination of the initial coating bond strength prediction model is associated with the Quantum Approximate Optimization Algorithm (QAOA) parameters of the quantum-enhanced intelligent optimizer, and the total loss function is used as the cost function of the QAOA parameters; In this embodiment, the hyperparameter combination includes network structure parameters, learning rate, total loss weight , total physical loss weight , Dropout ratio, etc. The hyperparameter optimization problem is mapped to the cost function minimization problem of QAOA, and each hyperparameter in the hyperparameter combination is mapped to a set of computational ground states. The formula of the cost function is:

[0048] In the formula, is the cost function, i.e., the cost Hamiltonian; is the hyperparameter combination; is the hyperparameter search space; is the hyperparameter combination The model constructed in the verification set uses the total loss value obtained by the total loss function; is the computational ground state; is the conjugate transpose of the computational ground state; A-42: The quantum-enhanced intelligent optimizer is initialized using Logistic chaotic mapping to generate initial QAOA parameters and enter a hybrid iterative optimization loop; The formula is:

[0049] In the formula, , is the first chaotic variable; n+ 1, n is the control parameter; compared with random initialization, chaotic initialization can ensure that the population is uniformly distributed in the solution space and enhance diversity; is the iteration index; n The generated chaotic sequence is linearly mapped to the typical range [0, 2π] of QAOA parameters:

[0050] In the formula, , is the initial QAOA parameter; is the initial cost parameter; is the initial mixing parameter; is the number of layers of QAOA; is the total number of layers of QAOA; , is the first chaotic variable; - 1, + - 1 chaotic variable; A-43: In each mixing iteration optimization cycle, run the QAOA circuit on the quantum computer using the current QAOA parameters to obtain the current quantum state, and sample the current hyperparameter combination; On the quantum computer, prepare the initial uniform superposition state and apply the parameterized quantum circuit The formula is:

[0051] In the formula, is the parameterized unitary operator, and is the overall evolution matrix of the QAOA circuit, which acts on the initial quantum state to generate a superposition state containing optimization problem information, and its behavior is completely controlled by the parameter ; is the QAOA parameter of the first mixing iteration; is the cost parameter of the first mixing iteration; is the mixing parameter of the first mixing iteration; is the number of mixing iterations; ​​​​

[0052] In the formula, It is a product of consecutive products; This is used for problem unit evolution, which encodes optimization problem information into quantum states; For hybrid unit evolution, used for mixing and exploration in the solution space; , For the first The first mixed iteration The QAOA parameters of the layer; For the first The first mixed iteration Cost parameters of the layer; For the first The first mixed iteration Layer blending parameters; Operators for hybrid Hamiltonians to achieve quantum tunneling and state transition; It is a quantum state; For cost Hamiltonians, the quantum representation of the optimization problem; i The imaginary unit; It is a natural constant; For quantum states Multiple measurements are performed to obtain a series of measurement results. Each measurement result is decoded to obtain a sampled hyperparameter combination. ,in, j This is a hyperparameter combination indicator; A-44: On a classic computer, using a sample dataset, quickly train alternative coating adhesion prediction models based on the current hyperparameter combination, and use a cost function to calculate the current cost value of the alternative coating adhesion prediction models. The formula is:

[0053] In the formula, For the first The expected cost of the next hybrid iteration, i.e., the cost value; Hyperparameter combination The constructed model on the validation set The total loss value obtained using the total loss function; For quantum states Number of measurements performed; A-45: Generate a flight random number. If the flight random number is less than the preset threshold, perform Levy flight exploration, update the current QAOA parameters, and obtain the updated QAOA parameters. Otherwise, perform gradient descent to update the current QAOA parameters, obtain the updated QAOA parameters, and return the hyperparameter combination acquisition and cost value calculation steps. Generate a uniformly distributed flight random number. r~ If the flight random number r Less than the preset threshold To conduct Levy flight exploration, the formula is:

[0054] In the formula, for Levy Distributed random numbers; b for Levy Step length, and b ∈[1,2]; For the first QAOA parameters for +1 mixed iterations; For the first Cost parameters for +1 mixed iteration; For the first +1 mixing iterations of mixing parameters; For adaptive step size; This is element-wise multiplication;

[0055] In the formula, This is the initial step size; These are the maximum and minimum values ​​of the convergence factor; This represents the maximum number of iterations. , To adjust the parameters; It is the hyperbolic tangent function; Otherwise, gradient descent is performed, and the cost gradient is estimated on a quantum computer using the parameter shift rule, as follows:

[0056] In the formula, For the expected cost versus parameters The gradient, i.e. ; The expected cost after a positive shift; The expected cost after negative shift;

[0057] In the formula, is the classical learning rate, and is the step size of gradient descent; Let be the gradient of the expected cost, a vector pointing in the direction of the steepest ascent of the cost function; A-46: If the current number of mixed iterations reaches the mixed iteration number threshold or the cost improvement is lower than the improvement threshold, terminate the mixed iteration optimization step and output the optimal hyperparameter combination corresponding to the optimal cost value; A-47: Set the hyperparameters of the initial coating adhesion force prediction model according to the optimal hyperparameter combination to obtain the final coating adhesion force prediction model.

[0058] In an optional implementation, the multi-modal non-destructive testing data is input into the coating adhesion force prediction model pre-trained based on the physical information multi-modal fusion algorithm for adhesion force prediction and integration to obtain a coating adhesion force spatial distribution map, including: S2021: Data preprocessing is performed on the multi-modal non-destructive testing data to obtain preprocessed multi-modal non-destructive testing data, and the preprocessed multi-modal non-destructive testing data is input into the coating adhesion force prediction model pre-trained based on the physical information multi-modal fusion algorithm; In this embodiment, the data preprocessing includes: Image data: Distortion correction, white balance adjustment, contrast limited adaptive histogram equalization enhancement are performed on the macroscopic and microscopic images, and the images are cropped to a uniform size (such as 224x224 pixels); Acoustic emission signal: The original waveform is subjected to 20kHz-2MHz band-pass filtering, and then key features such as event count, ring count, energy, amplitude, duration, rise time are extracted, or the processed time-domain waveform is directly input into the model; Thermal imaging map: Non-uniformity correction and noise filtering are performed on the original temperature sequence; then, the temperature-time curve of each pixel point is fitted, and key thermal features such as peak temperature, maximum temperature difference, and cooling rate constant are extracted; or the temperature sequence frames are directly stacked into a pseudo-three-dimensional data block to input into the model; S2022: The macroscopic image feature extractor of the coating adhesion force prediction model is used to extract the macroscopic feature map of the macroscopic image in the preprocessed multi-modal non-destructive testing data, and the macroscopic feature map is input into the global average pooling layer for global average pooling to obtain the macroscopic image feature (a 2048-dimensional feature vector); S2023: The microscopic image feature extractor of the coating adhesion force prediction model is used to extract the microscopic feature map of the microscopic image in the preprocessed multi-modal non-destructive testing data, and the microscopic feature map is input into the global average pooling layer for global average pooling to obtain the microscopic image feature (a 1280-dimensional feature vector); S2024: The acoustic time series feature extractor composed of 3 fully connected layers of the coating adhesion force prediction model is used, and ReLU activation function and Dropout layer are used after each layer to prevent overfitting, to extract the acoustic time series feature (a 256-dimensional feature vector) of the acoustic emission signal in the preprocessed multi-modal non-destructive testing data; S2025: The thermal imaging time sequence feature extractor using the coating adhesion force prediction model is composed of two 3D convolution layers, one 3D maximum pooling layer and one flattening layer, and is aimed at capturing the dynamic change pattern of temperature in time and space, and extracting the thermal imaging time sequence features (spatiotemporal dynamic features, 512-dimensional feature vectors) of the thermal imaging atlas in the preprocessed multi-modal non-destructive testing data; S2026: The macroscopic image features, microscopic image features, acoustic time sequence features and thermal imaging time sequence features are spliced to obtain a fusion feature vector, and input to the attention fusion module of the coating adhesion force prediction model. The module calculates the importance weight of each feature dimension to generate a weighted fusion feature vector, so that the model can focus on the most critical information and perform attention fusion to obtain a weighted fusion feature; S2027: According to the weighted fusion feature, the physical parameter output layer of the coating adhesion force prediction model is used to generate physical parameters (parameters with clear physical meaning); the physical parameters include interfacial thermal resistance, microcrack density and interfacial adhesion strength; In this embodiment, the physical parameter output layer includes a thermal resistance branch, a crack density branch and an adhesion strength branch; Thermal resistance branch: a fully connected layer outputs a scalar value, which is the predicted interfacial thermal resistance (unit: m²K / W); Crack density branch: a fully connected layer outputs a scalar value, which is the predicted microcrack density (unit: pieces / mm²); Adhesion strength branch: a fully connected layer outputs a scalar value, which is the predicted interfacial adhesion strength (unit: MPa); S2028: According to the physical parameters, the adhesion force prediction layer of the coating adhesion force prediction model is used to call the Softmax function to generate the adhesion force grade (output the probability distribution of belonging to the four grades of "excellent, good, medium and poor"), and combine the corresponding physical parameters to obtain the adhesion force prediction result; S2029: According to the spatial coordinate value of each detection point in the global coordinate system, the adhesion force prediction results of all detection points are integrated to obtain the corresponding coating adhesion force spatial distribution map; In this embodiment, the spatial coordinates of each detection point are associated with the corresponding adhesion force prediction results, and the data of these discrete points are expanded to the entire blade surface through an interpolation algorithm (such as Kriging interpolation) to generate a continuous and visual "blade coating adhesion force spatial distribution map".

[0059] In an alternative implementation, the coating adhesion force spatial distribution map is input to the high-fidelity digital twin body pre-constructed based on digital twin technology for coupling simulation and life analysis, to obtain a coating residual life distribution map, including: S2031: Import the three-dimensional point cloud model of the water turbine blade to be evaluated into simulation software (such as ANSYS or ABAQUS, etc.), and perform high-quality hexahedral dominant meshing on it, especially extremely fine meshing in the coating area with a thickness of only microns or millimeters, to ensure that the physical field at the interface can be accurately captured, and a corresponding three-dimensional simulation model is obtained; S2032: Based on digital twinning technology, define the elastic-plastic material model of the substrate and the coating in the three-dimensional simulation model, and set the corresponding damage evolution equation of the elastic-plastic material model of the coating, to obtain a pre-constructed high-fidelity digital twin; In this embodiment, the substrate: define the elastic-plastic constitutive model of the blade substrate material (such as stainless steel); Coating: define the elastic-plastic constitutive model of the coating, and introduce a damage evolution equation; for example, use the continuous medium damage mechanics model, and the relationship between the damage variable and the elastic modulus is When =0, the material is intact; when =1, the material is completely failed; Digital twin generation: the defined geometry, mesh, material model, and boundary condition setting together constitute a high-fidelity digital twin of the blade; S2033: Map the coating bonding force spatial distribution map as the initial condition of the damage evolution equation to the high-fidelity digital twin, and calibrate the high-fidelity digital twin; Take the "bonding force spatial distribution map" as the initial condition, map it to the high-fidelity digital twin coating part through scripts or APIs, and get the initial damage variable , establish a mapping relationship, and the formula is:

[0060] In the formula, is the initial damage variable; is the interfacial bonding strength; is the ultimate tensile strength of the coating or substrate material; is the shape coefficient, used to adjust the steepness of the curve; is a natural constant; Use the known failure cases in the historical operation data of the blade to calibrate the model, adjust the parameters in the damage evolution equation in reverse, make the simulation results consistent with the historical records, and improve the prediction accuracy of the model; S2034: Perform coupled simulation and damage accumulation on the high-fidelity digital twin under the simulation of a specific working condition, to obtain the damage value of each detection point; In this embodiment, the future power generation plan (such as the head, flow, and output curve of the next year) is obtained from the power station monitoring system; these data are taken as input conditions for the high-fidelity digital twin to perform fluid-structure coupling simulation; the simulation software calculates the water flow pressure field, which is applied to the structure finite element module to calculate the dynamic stress and strain field of the blade; After each time step (or load cycle), the damage increment of this step is calculated according to the stress level and the current damage value of each grid in the coating part of the high-fidelity digital twin ; The total damage value is calculated using the Miner linear cumulative damage theory , wherein, is the damage value of the i-th grid, l is the grid indicator, l is the number of cycles that lead to failure at the current stress level; and the damage value of each grid is updated ; S2035: If the damage value exceeds the damage threshold for the first time, the detection point is taken as a key detection point, and the coupling simulation time experienced by the key detection point is taken as the corresponding predicted remaining life; S2036: If the damage value does not exceed the damage threshold, the coupling simulation and damage accumulation are continued based on the detection point; S2037: The predicted remaining life of all key detection points is integrated to obtain a coating remaining life distribution map according to the spatial coordinate values of each key detection point in the global coordinate system.

[0061] In an optional implementation, a method for constructing an intelligent maintenance decision model includes: B-1: Taking the hydropower station operation and maintenance environment as a simulation environment; B-2: Defining the state space of the agent of the intelligent maintenance decision model, including global statistical features, local key area features, spatial context features, operation state vectors corresponding to hydropower station state data, logistics state vectors, and economic state vectors; The global statistical features of the force space distribution map include: Deterioration area ratio (for example, calculate the percentage of the area with a “poor” (level 3) bonding force in the total blade area. This is a very intuitive risk indicator), average bonding force / standard deviation (the average value reflects the overall level, and the standard deviation reflects the uniformity of damage, and a high standard deviation may mean that there is local severe damage), number of deteriorated areas (through connected component analysis, calculate the number of all deteriorated areas, and multiple small damages may be more risky than a single large damage); The global statistical features of the remaining life distribution map include: ​​Shortest remaining life (this is the most critical feature, directly determines the urgency of maintenance), mean remaining life / standard deviation (reflects the overall life expectancy and the unevenness of life distribution), critical life area ratio (for example, calculate the area ratio of the region with remaining life less than 3 months); Key areas include: weakest area (find the point (or area) with the worst bonding force in the bonding force spatial distribution map), shortest life area (find the point (or area) with the shortest remaining life in the remaining life distribution map), high-risk area (define a comprehensive risk index, find the area with the highest comprehensive risk index); The formula is:

[0062] In the formula, is the comprehensive risk index; is the normalization function; is the bonding force degradation value, which refers to the bonding force evaluation result of a certain area, which is the average value of the bonding force level of all grids in the area (for example, level 0-3, the larger the value, the worse); is the shortest remaining life of the area; , is the risk weight; Local key area features include: Position information: three-dimensional coordinates of the centroid of the area; This is very important because damage at different locations (such as the water inlet side vs. the water outlet side) has different maintenance strategies and risks; Size information: area or equivalent diameter of the area; Severity indicators: average bonding force, minimum remaining life in the area; Spatial context features include: Radial distribution features: divide the blade into multiple annular zones (such as 0-10%, 10%-20%, …, 90%-100%) along the radial direction; Calculate the average bonding force and average remaining life in each annular zone; This can reveal whether the damage is closer to the blade tip (where the stress is usually higher); Spatial autocorrelation: calculate the Moran's I index to measure whether the damage is clustered, random, or discrete; A clustered damage pattern may indicate a common cause (such as cavitation); Operation state vector includes: current time, current head, current flow, current output, predicted load curve for a period of time in the future, etc.; Logistics state vector includes: spare parts inventory (such as the amount of repair coating material), maintenance team availability (such as whether they are at the plant, whether they have other tasks), planned downtime window, etc.; Economic state vector includes: current electricity price, expected future electricity price, unit downtime loss cost, unit maintenance cost, etc.; B-2: defining the action space of the agent of the intelligent maintenance decision model, including: {immediate overhaul, immediate local repair, delay for 1 month, delay for 3 months, 10% load reduction operation, 20% load reduction operation, no operation}; B-3: defining a complex reward function for evaluating the long-term benefits of each action;

[0063] wherein, is the reward value; is the maintenance cost; is the unit time downtime loss; is the maintenance duration; is the power generation loss during downtime; is the potential loss penalty caused by risks (such as peeling during operation); B-4: training the agent for millions of interactions in a simulated environment, building the intelligent maintenance decision model, and setting up an experience replay pool; The agent learns an optimal strategy through continuous trial and error, which can select the action with the highest expected cumulative reward in any given state.

[0064] In an optional implementation, the coating adhesion force spatial distribution map and the coating residual life distribution map are input into the intelligent maintenance decision model pre-constructed based on the reinforcement learning algorithm to generate decisions, and a blade maintenance decision is obtained, including: S2041: extracting global statistical features, local key area features, and spatial context features of the coating adhesion force spatial distribution map and the coating residual life distribution map, and combining with the state data of the hydropower station to obtain a complete state observation; S2042: writing the complete state observation into the state space of the agent of the intelligent maintenance decision model pre-constructed based on the reinforcement learning algorithm; S2043: based on the state space, using the agent to select the optimal action in the preset action space to obtain the blade maintenance decision, and collecting real-time experience in the decision generation process; S2044: randomly extracting a number of historical experiences in the experience replay pool, and combining the real-time experience to continuously train the intelligent maintenance decision model to obtain an updated intelligent maintenance decision model.

[0065] In an optional implementation, the coating adhesion force spatial distribution map and the coating residual life distribution map are visualized using a high-fidelity digital twin, and a blade coating adhesion force evaluation report is generated in combination with the blade maintenance decision, including: S2051: loading the high-fidelity digital twin of the water turbine blade to be evaluated; S2052: The coating adhesion spatial distribution map and the coating residual life distribution map are superimposed and rendered on the high-fidelity digital twin in the form of heat maps of different colors as independent layers, and visualized and displayed; S2053: The coating adhesion spatial distribution map, the coating residual life distribution map, and the corresponding blade maintenance decision are written into a preset report template to generate a blade coating adhesion evaluation report, and visualized and displayed; In this embodiment, the blade coating adhesion evaluation report is a structured PDF or web page report, which can include: Executive summary: overall health score of the blade, number and location of high-risk areas, predicted earliest failure time, and blade maintenance decision; Detailed analysis chart: adhesion distribution histogram, life prediction trend chart, and cost-benefit comparison chart of different blade maintenance decisions; Decision basis: detailed explanation of why the intelligent decision model recommends the maintenance strategy, listing the risk analysis and economic calculation behind it; Appendix: contains links to the original data of all detection points for experts to conduct in-depth review.

[0066] The embodiment of the present application also provides a water turbine blade coating adhesion evaluation device, which refers to Figure 3 , shows a functional unit diagram of a water turbine blade coating adhesion evaluation device 300, which can include the following units: A multi-modal data acquisition unit 301 is used to synchronously acquire multi-modal non-destructive testing data of the coating area of the water turbine blade to be evaluated; A coating adhesion prediction unit 302 is used to input the multi-modal non-destructive testing data into a coating adhesion prediction model based on a physical information multi-modal fusion algorithm to predict and integrate the adhesion, and obtain a coating adhesion spatial distribution map; A coating residual life analysis unit 303 is used to input the coating adhesion spatial distribution map as an initial condition into a high-fidelity digital twin based on digital twin technology to perform coupling simulation and life analysis, and obtain a coating residual life distribution map; A blade maintenance decision generation unit 304 is used to input the coating adhesion spatial distribution map and the coating residual life distribution map into an intelligent maintenance decision model based on a reinforcement learning algorithm to generate a decision, and obtain a blade maintenance decision; A visual display unit 305 is used to use a high-fidelity digital twin to visually display the coating adhesion spatial distribution map and the coating residual life distribution map, and generate a blade coating adhesion evaluation report in combination with the blade maintenance decision.

[0067] Based on the same inventive concept, another embodiment of the present application provides an electronic device, comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory complete communication with each other through the communication bus, the memory is used for storing a computer program; the processor is used for executing the program stored on the memory to realize the water turbine blade coating bonding force evaluation method.

[0068] The communication bus mentioned above can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EI) bus, etc. The communication bus can be divided into an address bus, a data bus, a control bus, etc. For the convenience of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus. The communication interface is used for communication between the terminal and other devices. The memory can include a Random Access Memory (RAM) and can also include a non-volatile memory, for example, at least one disk memory. Optionally, the memory can also be at least one storage device located away from the aforementioned processor.

[0069] The processor mentioned above can be a general-purpose processor, including a Central Processing Unit (CPU), a Network Processor (NP), etc. It can also be a Digital Signal Processing (DSP), an Application Specific Integrated Circuit (ASIC), a Field-Programmable Gate Array (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0070] In addition, to achieve the above-mentioned purpose, an embodiment of the present application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the water turbine blade coating bonding force evaluation method of the embodiment of the present application.

[0071] Those skilled in the art will appreciate that embodiments of the present application can be readily used as a method, apparatus, or computer program product. Accordingly, embodiments of the present application can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present application can take the form of a computer program product on one or more computer-usable hardware storage devices (including, but not limited to, disk storage, CD-ROMs, optical storage devices, etc.) embodying computer readable program code.

[0072] Embodiments of the present application are described herein with reference to the Figure 1 one or more functions specified in a flow or multiple flows and / or blocks. Figure 1 means for performing one or more functions specified in a flow or multiple flows and / or blocks.

[0073] These computer program instructions can also be stored in a computer- readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instructions which implement the Figure 1 one or more functions specified in a flow or multiple flows and / or blocks. Figure 1 means for performing one or more functions specified in a flow or multiple flows and / or blocks.

[0074] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the Figure 1 one or more functions specified in a flow or multiple flows and / or blocks. Figure 1 means for performing one or more functions specified in a flow or multiple flows and / or blocks.

[0075] The above, merely, is a specific implementation of the present application, but the protection scope of the present application is not limited to this, any skilled person in the art can easily think of various equivalent modifications or replacements within the technical range disclosed by the present application, these modifications or replacements should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A method for evaluating the adhesion of a coating on a water turbine blade, characterized by, The method comprises: Synchronously collecting multi-modal non-destructive testing data of the coating area of the water turbine blade to be evaluated; the multi-modal non-destructive testing data comprises macroscopic images, microscopic images, acoustic emission signals and thermal imaging maps; Inputting the multi-modal non-destructive testing data into a coating adhesion force prediction model pre-trained based on a physical information multi-modal fusion algorithm to perform adhesion force prediction and integration, and obtaining a coating adhesion force spatial distribution map; Inputting the coating adhesion force spatial distribution map as an initial condition into a high-fidelity digital twin pre-constructed based on a digital twin technology to perform coupling simulation and life analysis, and obtaining a coating residual life distribution map; Inputting the coating adhesion force spatial distribution map and the coating residual life distribution map into an intelligent maintenance decision model pre-constructed based on a reinforcement learning algorithm to generate a decision, and obtaining a blade maintenance decision; Using the high-fidelity digital twin, visualizing the coating adhesion force spatial distribution map and the coating residual life distribution map, and combining the blade maintenance decision to generate a blade coating adhesion force evaluation report.

2. The method for evaluating the bonding force of a water turbine blade coating according to claim 1, wherein Synchronously collecting multi-modal non-destructive testing data of the coating area of the water turbine blade to be evaluated, comprising: Based on a laser scanning technology, acquiring a three-dimensional point cloud model of the water turbine blade to be evaluated, and establishing a global coordinate system of the water turbine blade; Planning a scanning path covering the coating area of the entire blade surface on the three-dimensional point cloud model, and setting a plurality of detection points on the blade surface of the water turbine blade according to the scanning path; Spatially registering the multi-modal data acquisition probes in the global coordinate system to obtain spatially registered multi-modal data acquisition probes; Using the multi-modal data acquisition probes, synchronously collecting multi-modal data at each detection point to obtain multi-modal non-destructive testing data of the coating area of the water turbine blade to be evaluated.

3. The method of claim 2, wherein the step of determining the adhesion of the coating layer is performed by measuring the adhesion of the coating layer on the test piece. The coating adhesion force prediction model comprises an image feature extraction branch and a time series feature extraction branch connected in parallel, and an attention fusion module, a physical parameter output layer and an adhesion force prediction layer connected in sequence; The image feature extraction branch comprises a macroscopic image feature extractor based on a ResNet-50 algorithm and a microscopic image feature extractor based on an EfficientNet-B0 algorithm connected in parallel, and a global average pooling layer; The time series feature extraction branch comprises an acoustic time series feature extractor based on an MLP algorithm and a thermal imaging time series feature extractor based on a 3D-CNN algorithm connected in parallel; The coating adhesion force prediction model is provided with a total loss function combining physical information and data loss, the total loss function comprises a total physical loss function and a data loss function, the total physical loss function comprises a heat conduction physical loss function, a fatigue damage physical loss function and a physical consistency loss function, and the formula is: wherein is a total loss function; is a data loss function; is a total physical loss function; , is a total loss weight; wherein, is a heat conduction physics loss function; is a fatigue damage physics loss function; is a physics consistency loss function; , , is a total physics loss weight; wherein is the MSE function; is the predicted surface temperature response curve calculated from the interface thermal resistance output by the physical parameter output layer; is the measured surface temperature response curve; wherein is a function for calculating the acoustic emission signal energy; is the acoustic emission signal energy; is a mapping function for mapping the stress factor to the damage space; is the stress intensity factor range; wherein is the interfacial thermal resistance; is the microcrack density; is the base interfacial thermal resistance; is the damage-thermal resistance coupling coefficient; The intelligent maintenance decision model is constructed based on a PPO algorithm, and the intelligent maintenance decision model is provided with an agent and an experience replay pool.

4. The method for evaluating the bonding force of a water turbine blade coating according to claim 3, characterized by, The construction method of the coating adhesion force prediction model comprises: Preparing a plurality of coating test blocks of the water turbine blade simulating different working conditions in a laboratory, collecting sample multi-modal non-destructive testing data of each coating test block; For each coating test block, destructive testing and surface temperature testing are performed to obtain corresponding bond strength true value labels and measured surface temperature response curves, and are added to corresponding sample multi-modal non-destructive testing data to obtain a sample data set; Based on the physical information multi-modal fusion algorithm, an initial coating bond strength prediction model is constructed; Based on the total loss function, the hyperparameters of the initial coating bond strength prediction model are optimized using a quantum-enhanced intelligent optimizer to obtain a final coating bond strength prediction model.

5. The method for evaluating the bonding force of a water turbine blade coating according to claim 4, characterized by, Based on the total loss function, the hyperparameters of the initial coating bond strength prediction model are optimized using a quantum-enhanced intelligent optimizer to obtain a final coating bond strength prediction model, including: Combining the hyperparameters of the initial coating bond strength prediction model to the QAOA parameters of the quantum-enhanced intelligent optimizer, and taking the total loss function as the cost function of the QAOA parameters; Initialize the quantum-enhanced intelligent optimizer using the Logistic chaotic mapping to generate initial QAOA parameters and enter the hybrid iterative optimization loop; In each hybrid iterative optimization loop, run the QAOA circuit on the quantum computer using the current QAOA parameters to obtain the current quantum state, and sample the current hyperparameter combination; On the classical computer, use the sample data set to quickly train a candidate coating bond strength prediction model based on the current hyperparameter combination, and use the cost function to calculate the current cost value of the candidate coating bond strength prediction model; Generate a flight random number, if the flight random number is less than a predetermined threshold, perform Levy flight exploration to update the current QAOA parameters to obtain updated QAOA parameters, otherwise, perform gradient descent utilization to update the current QAOA parameters to obtain updated QAOA parameters, and return to the hyperparameter combination obtaining and cost value calculating step; If the current number of hybrid iterations reaches a hybrid iteration number threshold or the cost value improvement is lower than an improvement threshold, terminate the hybrid iterative optimization step and output the optimal hyperparameter combination corresponding to the optimal cost value; According to the optimal hyperparameter combination, set the hyperparameters of the initial coating bond strength prediction model to obtain the final coating bond strength prediction model.

6. The method for evaluating the bonding force of a water turbine blade coating according to claim 5, wherein Input the multi-modal non-destructive testing data into the coating bond strength prediction model based on the physical information multi-modal fusion algorithm for bond strength prediction and integration to obtain a coating bond strength spatial distribution map, including: Data preprocessing is performed on the multi-modal non-destructive testing data to obtain preprocessed multi-modal non-destructive testing data, which is input into the coating bond strength prediction model based on the physical information multi-modal fusion algorithm; Using the macro image feature extractor of the coating bond strength prediction model, the macro feature map of the macro image in the preprocessed multi-modal non-destructive testing data is extracted and input into the global average pooling layer for global average pooling to obtain the macro image feature; Using the macro image feature extractor of the coating bond strength prediction model, the macro feature map of the macro image in the preprocessed multi-modal non-destructive testing data is extracted and input into the global average pooling layer for global average pooling to obtain the macro image feature; an acoustic time sequence feature extractor using the coating adhesion force prediction model, extracts acoustic time sequence features of acoustic emission signals in the preprocessed multi-modal non-destructive testing data; a thermal imaging time sequence feature extractor using the coating adhesion force prediction model, extracts thermal imaging time sequence features of thermal imaging maps in the preprocessed multi-modal non-destructive testing data; the macroscopic image features, the microscopic image features, the acoustic time sequence features and the thermal imaging time sequence features are spliced to obtain a fusion feature vector, and the fusion feature vector is input into an attention fusion module of the coating adhesion force prediction model to perform attention fusion to obtain a weighted fusion feature; a physical parameter output layer of the coating adhesion force prediction model is used according to the weighted fusion feature to generate a physical parameter; the physical parameter includes an interfacial thermal resistance, a micro-crack density and an interfacial adhesion strength; a combination force prediction layer of the coating adhesion force prediction model is used according to the physical parameter to call a Softmax function to generate a combination force grade, and a combination force prediction result is obtained in combination with the corresponding physical parameter; the combination force prediction results of all detection points are integrated according to the spatial coordinate values of each detection point in the global coordinate system to obtain a coating adhesion force spatial distribution map.

7. The method for evaluating the bonding force of a water turbine blade coating according to claim 6, wherein The coating adhesion force spatial distribution map is input into a high-fidelity digital twin body pre-constructed based on a digital twin technology to perform coupling simulation and life analysis, and a coating residual life distribution map is obtained, including: a three-dimensional point cloud model of the water turbine blade to be evaluated is imported into a simulation software, and a high-quality hexahedral dominant mesh is divided to obtain a corresponding three-dimensional simulation model; based on the digital twin technology, an elastic-plastic material model of the matrix and the coating is defined in the three-dimensional simulation model, and a damage evolution equation corresponding to the elastic-plastic material model of the coating is set to obtain a pre-constructed high-fidelity digital twin body; the coating adhesion force spatial distribution map is mapped to the high-fidelity digital twin body as an initial condition of the damage evolution equation, and the high-fidelity digital twin body is calibrated; the high-fidelity digital twin body is subjected to coupling simulation and damage accumulation under simulated specific working conditions to obtain a damage value of each detection point; if the damage value exceeds the damage threshold for the first time, the detection point is taken as a key detection point, and a coupling simulation time experienced by the key detection point is taken as a corresponding predicted residual life; if the damage value does not exceed the damage threshold, the coupling simulation and damage accumulation are continued based on the detection point; the predicted residual lives of all key detection points are integrated according to the spatial coordinate values of each key detection point in the global coordinate system to obtain a coating residual life distribution map.

8. The method for evaluating the bonding force of a water turbine blade coating according to claim 7, wherein The coating adhesion force spatial distribution map and the coating residual life distribution map are input into an intelligent maintenance decision model pre-constructed based on a reinforcement learning algorithm to generate a decision, and a blade maintenance decision is obtained, including: global statistical features, local key area features and spatial context features of the coating adhesion force spatial distribution map and the coating residual life distribution map are extracted, and state data of a hydropower station is obtained to obtain a complete state observation; the complete state observation is written into a state space of an agent of the intelligent maintenance decision model pre-constructed based on the reinforcement learning algorithm; Based on the state space, an agent is used to select the optimal action in the preset action space to obtain a blade maintenance decision, and real-time experience in the decision generation process is collected; A number of historical experiences are randomly extracted from the experience replay pool, and the intelligent maintenance decision model is continuously trained in combination with real-time experience to obtain an updated intelligent maintenance decision model.

9. The method of claim 8, wherein the step of determining the adhesion of the coating is performed by measuring the peel strength of the coating. The coating adhesion force spatial distribution map and the coating residual life distribution map are visualized using a high-fidelity digital twin, and a blade coating adhesion force evaluation report is generated in combination with the blade maintenance decision, including: Load the high-fidelity digital twin of the water turbine blade to be evaluated; The coating adhesion force spatial distribution map and the coating residual life distribution map are superimposed and rendered on the high-fidelity digital twin in the form of heat maps of different colors as independent layers, and are visualized; The coating adhesion force spatial distribution map, the coating residual life distribution map, and the corresponding blade maintenance decision are written into a preset report template to generate a blade coating adhesion force evaluation report and visualized.

10. A device for evaluating the adhesion of a coating on a water turbine blade, for implementing the method for evaluating the adhesion of a coating on a water turbine blade according to any one of claims 1 to 9, characterized in that, The device comprises: A multi-modal data acquisition unit for synchronously acquiring multi-modal non-destructive testing data of the coating area of the water turbine blade to be evaluated; A coating adhesion force prediction unit for inputting the multi-modal non-destructive testing data into a coating adhesion force prediction model pre-trained based on a physical information multi-modal fusion algorithm for adhesion force prediction and integration to obtain a coating adhesion force spatial distribution map; A coating residual life analysis unit for inputting the coating adhesion force spatial distribution map as an initial condition into a high-fidelity digital twin pre-constructed based on digital twin technology for coupling simulation and life analysis to obtain a coating residual life distribution map; A blade maintenance decision generation unit for inputting the coating adhesion force spatial distribution map and the coating residual life distribution map into an intelligent maintenance decision model pre-constructed based on a reinforcement learning algorithm for decision generation to obtain a blade maintenance decision; A visual display unit for visualizing the coating adhesion force spatial distribution map and the coating residual life distribution map using a high-fidelity digital twin, and generating a blade coating adhesion force evaluation report in combination with the blade maintenance decision.

Images