Degradation simulation test system and evaluation method for whole life cycle of dredge pump

Abstract

The invention relates to the technical field of engineering equipment operation reliability testing and life evaluation, in particular to a degradation simulation test system and evaluation method for the whole life cycle of a dredge pump, which comprises a plurality of loading units capable of being independently controlled and a test evaluation unit, the test evaluation unit corrects the degradation model according to the real-time physical field data and the measured data, adjusts the loading combination mode of the plurality of loading units and reloads the physical field according to the error conditions of different degradation operation modules in the degradation model correction process, and repeats the process to obtain a corrected degradation model; and evaluating the service life of the dredge pump based on the degradation model. According to the invention, flexible loading of multi-factor combination is realized through cooperation of innovative degradation model design and a plurality of loading units which can be independently controlled, degradation data is accurately obtained, and the problem that a dredge pump degradation simulation test cannot be carried out by a dredge pump test platform based on multi-physics field coupling in the prior art is solved.

Description

Technical Field

[0001] This invention relates to the field of reliability testing and life evaluation technology for engineering equipment, and in particular to a degradation simulation test system and evaluation method for mud pumps throughout their entire life cycle. Background Technology

[0002] As the core component of a high-concentration sediment transport system, the mud pump undergoes complex working conditions, structural fatigue, and wear evolution during long-term continuous operation, and its performance degradation exhibits obvious multi-factor coupling characteristics.

[0003] Especially in practical engineering, the lifespan of mud pumps is affected by the following key factors: long-term abrasive action of sand-containing media, accelerated aging effect of environmental temperature, humidity and vibration loads, and sudden damage caused by cavitation impact under unsteady conditions. However, existing test platforms mostly focus on single-factor loading (such as sand-containing wear test benches) and lack multi-source degradation loading systems that can reproduce real service conditions, making it impossible to systematically obtain performance evolution data of mud pumps throughout their entire life cycle.

[0004] Therefore, there is a need for a degradation simulation test system for the entire life cycle of mud pumps that can simultaneously simulate the coupling effects of multiple physical fields, so as to provide support for the analysis and prediction model of mud pump life cycle. Summary of the Invention

[0005] Therefore, the present invention provides a degradation simulation test system and evaluation method for the entire life cycle of mud pumps, in order to solve the problem that the existing mud pump test platform cannot conduct mud pump degradation simulation tests based on multi-physics field coupling.

[0006] This invention provides a degradation simulation test system for the entire life cycle of a mud pump, comprising multiple independently controllable loading units and a test evaluation unit. Each loading unit is used to apply a physical field of a degradation factor to the mud pump, and the test evaluation unit is used to perform the following steps: Step 1: Establish an initial degradation model. The degradation model includes multiple degradation operation modules. Each degradation operation module corresponds to the degradation effect of one degradation factor or multiple degradation factors coupled together. The degradation model is used to input mud pump status data and physical field data of each degradation factor, and outputs predicted data of mud pump life quantitative indicators after collaborative operation of multiple degradation operation modules. Step 2: After the loading unit loads the physical field, obtain the real-time physical field data of the loading unit and the measured data of the mud pump life quantification index; Step 3: Correct the degradation model based on real-time physics data and measured data, and adjust the loading combination mode of multiple loading units and reload the physics field based on the error of different degradation operation modules during the degradation model correction process. Step 4: Repeat steps 2 and 3 to obtain the corrected degradation model, and evaluate the life of the mud pump based on the degradation model.

[0007] In a preferred embodiment: the degradation model is corrected based on real-time physics data and measured data, and the loading combination mode of multiple loading units is adjusted and the physics field is reloaded based on the error of different degradation operation modules during the degradation model correction process, including: The degradation calculation module in the degradation model is corrected based on real-time physical field data and measured data; Based on the correction amount of each degradation operation module, the error-dominant module is obtained; Based on the type of degradation effect corresponding to the error-dominant module, adjust the loading combination mode of multiple loading units and reload the physical field.

[0008] In a preferred embodiment: the degradation model includes multiple coupled operation layers, and multiple degradation operation modules are distributed in multiple coupled operation layers. The number of coupled degradation factors represented by each coupled operation layer increases progressively. Degradation operation modules representing the same degradation factor coupling relationship are connected between adjacent coupled operation layers, and the output data of the degradation operation module of the previous layer is used as the input data of the degradation operation module of the next connected layer. The degradation operation module of the first layer is used to input the physical field data and mud pump state data of its corresponding type, respectively, and the degradation operation module of the last layer is used to output the prediction data.

[0009] In a preferred embodiment: the degradation factors include abrasion, aging, and cavitation; the multiple coupled computation layers include a single-factor computation layer, a two-factor coupling layer, and a three-factor coupling layer. The degradation computation module in the single-factor computation layer is used to input the physical field data and mud pump status data of its corresponding type and output a first intermediate vector. The degradation computation module in the two-factor coupling layer is used to input the first intermediate vector corresponding to the coupled degradation factor it represents and output a second intermediate vector. The three-factor coupling layer is used to input all the second intermediate vectors and output the predicted data.

[0010] In a preferred embodiment: the degradation calculation module in the degradation model is corrected based on real-time physical field data and measured data, including: Based on the degradation model, the computational path of the error-dominant module is obtained; Freeze other degenerate operation modules outside the operation path in the degenerate model; Based on real-time physical field data and measured data, the degradation calculation module in the calculation path is corrected.

[0011] In a preferred embodiment: based on real-time physical field data and measured data, the degenerate computation module in the computation path is corrected, including: Real-time physical field data is input into the degradation model to obtain the predicted data output by the degradation model; The error function is obtained based on the difference between the predicted data and the measured data; Based on the error function, the degenerate computation modules in the computation path are corrected by gradient descent.

[0012] In a preferred embodiment: based on the correction amount of each degenerate operation module, the error-dominant module is obtained, including: The sum of the adjustments to each parameter in each degradation operation module is calculated and used as the correction amount for each degradation operation module; The degradation operation module with the highest correction value is selected as the error-dominant module.

[0013] In a preferred embodiment: an initial degradation model is established, including: Randomly set the parameters in the degradation model; By individually loading physical fields onto each loading unit, and based on the real-time physical field data of the loading unit and the measured data of the mud pump life quantification index after individually loading the physical fields, the random parameters in the degradation model are corrected to obtain the initial degradation model.

[0014] In a preferred embodiment: degradation factors include abrasion, aging and cavitation; the loading unit includes an automatic sand and water supply unit, a temperature, humidity and vibration integrated loading chamber, and a controllable cavitation induction unit; the quantification indicators of mud pump life include at least one of head retention rate, shaft power fluctuation coefficient, acoustic emission spectrum characteristics, vibration amplitude change rate, and flow efficiency.

[0015] This invention also provides a degradation simulation test evaluation method for the entire life cycle of mud pumps, applicable to any of the aforementioned degradation simulation test systems for the entire life cycle of mud pumps, comprising: Step 1: Establish an initial degradation model. The degradation model includes multiple degradation operation modules. Each degradation operation module corresponds to the degradation effect of one degradation factor or multiple degradation factors coupled together. The degradation model is used to input mud pump status data and physical field data of each degradation factor, and outputs predicted data of mud pump life quantitative indicators after collaborative operation of multiple degradation operation modules. Step 2: After the loading unit loads the physical field, obtain the real-time physical field data of the loading unit and the measured data of the mud pump life quantification index; Step 3: Correct the degradation model based on real-time physics data and measured data, and adjust the loading combination mode of multiple loading units and reload the physics field based on the error of different degradation operation modules during the degradation model correction process. Step 4: Repeat steps 2 and 3 to obtain the corrected degradation model, and evaluate the life of the mud pump based on the degradation model.

[0016] The beneficial effects of adopting the above scheme are: This invention provides a degradation simulation test system for the entire life cycle of mud pumps, comprising multiple independently controllable loading units and a test evaluation unit. Each loading unit is used to load a physical field of a degradation factor onto the mud pump. The test evaluation unit is used to establish an initial degradation model, correct the degradation model based on real-time physical field data from the loading units and measured data of the mud pump's life quantification index, and adjust the loading combination mode of multiple loading units and reload the physical field according to the error of different degradation calculation modules during the degradation model correction process. This process is repeated to obtain a corrected degradation model, and the life of the mud pump is evaluated based on the degradation model. The degradation model includes multiple degradation calculation modules, each corresponding to the degradation effect of one or more coupled degradation factors. The mud pump status data and physical field data are output as predicted data of the mud pump's life quantification index after collaborative calculation by multiple degradation calculation modules. This invention, through innovative degradation model design, decouples the independent effects of single factors and the coupled effects of multiple factors into multiple degradation calculation modules. These modules, combined with multiple independently controllable loading units, enable flexible loading of multiple factor combinations. A closed-loop mechanism, employing measured data to correct the model, model error-driven loading combination adjustment, and targeted supplementary data for further correction, locates the source of model error and dynamically switches loading unit combinations. This eliminates the interference from multiple physical fields, accurately acquires degradation data, and ultimately obtains a model that accurately characterizes the degradation laws of multi-factor coupling. This provides experimental support for the performance evolution analysis and life prediction of mud pumps throughout their entire life cycle, combining physical realism and model reliability. It solves the problem that existing mud pump test platforms cannot conduct mud pump degradation simulation tests based on multi-physical field coupling. Attached Figure Description

[0017] Figure 1 The system architecture diagram of the degradation simulation test system for the entire life cycle of mud pumps provided by the present invention; Figure 2 A flowchart of the degradation simulation test method for the entire life cycle of mud pumps provided by the present invention; Figure 3 for Figure 2 A detailed step diagram of step S203 is shown below; Figure 4 for Figure 3 A detailed step diagram of step S301 is shown below; Figure 5 This is a schematic diagram of the structure of a degradation model in one embodiment of the present invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] Combination Figure 1 and Figure 2 As shown, a specific embodiment of the present invention discloses a degradation simulation test system for the entire life cycle of a mud pump, including multiple independently controllable loading units 110 and a test evaluation unit 120. Each loading unit is used to load a physical field of a degradation factor onto the mud pump, and the test evaluation unit is used to perform the following steps: S201. Establish an initial degradation model. The degradation model includes multiple degradation operation modules. Each degradation operation module corresponds to one degradation factor or the degradation effect of multiple degradation factors coupled together. The degradation model is used to input mud pump status data and physical field data of each degradation factor, and outputs predicted data of mud pump life quantitative index after collaborative operation of multiple degradation operation modules. S202. After the loading unit loads the physical field, obtain the real-time physical field data of the loading unit and the measured data of the mud pump life quantification index. S203. Correct the degradation model based on real-time physical field data and measured data, and adjust the loading combination mode of multiple loading units and reload the physical field based on the error of different degradation operation modules during the degradation model correction process. S204. Repeat steps two and three to obtain the corrected degradation model, and evaluate the life of the mud pump based on the degradation model.

[0020] The system in this invention loads different degradation factor physical fields through independently controlled loading units to achieve multi-factor coupled simulation experiments of mud pumps. Its specific implementation can be achieved by superimposing any existing test platform. For example, in a specific embodiment, the degradation simulation test system for the entire life cycle of mud pumps is composed of a modular test platform. The platform includes a variable frequency drive main pump module for simulating different speed and load conditions, vulnerable blocks designed for key components such as impellers and pump casings, replaceable wear target material components for controllable wear loading and periodic replacement, abrasive fluid with adjustable concentration, an automatic sand and water supply system with particle transport stability control, a temperature, humidity and vibration integrated loading chamber that simulates common thermal cycles, vibration disturbances and high humidity environments in marine equipment, supports fixed frequency vibration and periodic temperature rise, and a controllable cavitation induction unit that generates controllable cavitation impact under specific pressure and speed conditions to evaluate the impact of extreme conditions on structural performance.

[0021] In the above description, the automatic sand and water supply unit, the temperature, humidity, and vibration integrated loading chamber, and the controllable cavitation induction unit are each used as a loading unit to apply the physical fields of the three degradation factors: abrasion, aging, and cavitation. This establishes a parallel loading path for the three factors of "abrasion-aging-cavitation," with each channel possessing independent control capabilities. It supports scheduled or random operating conditions to simulate the multi-factor collaborative degradation process in real-world applications. For example, when some loading units are operating, other units can be paused or operate under preset standard conditions, enabling arbitrary combination loading control.

[0022] However, in practice, when simulating physical fields with multiple influencing factors simultaneously, the loading paths of each factor often exhibit stronger coupling than actually occurs due to limitations such as the space of the experimental platform. For example, the flow of sand-containing media (abrasion) can alter the pressure distribution in the flow channel, affecting the critical conditions for cavitation; environmental temperature and humidity (aging) may cause changes in the elastic modulus of the material, indirectly affecting the transmission efficiency of vibration loads; and vibration loads (aging) may exacerbate the micro-cutting effect of abrasive particles on the surface. This results in a discrepancy between the actual load on the tested mud pump and the actual loading data from the loading unit. Therefore, this invention further adds an experimental evaluation unit. By establishing a hierarchical degradation model that integrates the independent effects of single factors and the coupling effects of multiple factors, the error is calculated by comparing it with measured indicators. Then, the loading combination of influencing factors is dynamically adjusted based on the error, thereby dynamically decoupling strong coupling interference and forcing the model to learn the true independent effects and controllable coupling laws of each factor. For example, if the module error corresponding to the abrasion calculation is significantly high, it indicates that the model's characterization error of the abrasion effect is large. In this case, the experimental evaluation unit instructs the automatic sand and water supply system to load the abrasion factor separately (pausing the temperature and humidity vibration chamber and the cavitation unit or setting it to standard operating conditions), collects "pure abrasion" data, and updates the model accordingly. As another example, if the module error corresponding to abrasion and aging is large, the sand and water supply system and the temperature and humidity vibration chamber are instructed to load collaboratively, collect dual-factor coupled data, and correct the corresponding coupled module. Ultimately, through a closed-loop iteration process—from addressing positioning errors to hardware switching and loading, and then to supplementary data correction—the parameter decoupling problem caused by physical field crosstalk in the experimental platform was solved. This provides experimental support for the analysis of mud pump lifespan patterns that is both physically realistic and has model credibility. It should be noted that the experimental evaluation unit described above can be implemented using any existing hardware with computational and control capabilities. The selection of mud pump lifespan quantification indicators can also be flexibly designed according to actual conditions, as long as they can be used to evaluate mud pump lifespan. These indicators include, but are not limited to, head retention rate, shaft power fluctuation coefficient, acoustic emission spectrum characteristics, vibration amplitude change rate, and flow efficiency.

[0023] Furthermore, in combination Figure 3In a preferred embodiment, step S203, which involves correcting the degradation model based on real-time physical field data and measured data, and adjusting the loading combination mode of multiple loading units and reloading the physical field based on the error of different degradation operation modules during the degradation model correction process, specifically includes: S301. Correct the degradation calculation module in the degradation model based on real-time physical field data and measured data; S302. Based on the correction amount of each degradation operation module, obtain the error-dominant module; S303. Based on the type of degradation effect corresponding to the error-dominant module, adjust the loading combination mode of multiple loading units and reload the physical field.

[0024] In this embodiment, the degradation model is corrected by modifying the degradation operation module. Since each degradation operation module in this embodiment corresponds to a degradation factor or the degradation effect of multiple degradation factors coupled together, this embodiment directly locks the "error-dominant module" that contributes the most to the current error. Based on the type of degradation effect corresponding to the error-dominant module (such as single-factor abrasion, two-factor abrasion-aging coupling, etc.), the combination mode of the loading unit is adjusted in a targeted manner. If the error originates from a single-factor module (such as abrasion), only the corresponding abrasion loading unit (automatic sand and water supply system) is activated while other units are kept in standard working conditions. "Pure single-factor" data is collected to strengthen the model's representation of this independent effect. If the error originates from a coupling module (such as abrasion-aging coupling layer), the corresponding two-factor loading unit (sand and water supply + temperature and humidity vibration chamber) is activated in a coordinated manner to focus on supplementing coupling data to correct the perception of interaction effects. This closed-loop loading mechanism, which traces errors to their source and controls them precisely, significantly reduces experimental costs and time consumption by minimizing ineffective loading combinations (such as temporarily avoiding irrelevant factors with high errors). At the same time, it ensures the model's learning accuracy for each degradation effect (independent / coupled), ultimately enabling the degradation model to more realistically reflect the performance evolution of mud pumps under the synergistic effect of multiple factors.

[0025] The selection method for the error-dominant module can be flexibly designed according to actual conditions. In a preferred embodiment, the above step: S302, obtaining the error-dominant module based on the correction amount of each degradation operation module, specifically includes: The sum of the adjustments to each parameter in each degradation operation module is calculated and used as the correction amount for each degradation operation module; The degradation operation module with the highest correction value is selected as the error-dominant module.

[0026] Furthermore, since the degradation model in this embodiment needs to be dynamically updated based on the error situation, the model initialization can start with single-factor loading to obtain "pure" single-factor influence data and establish the basic degradation understanding of the model, thereby accelerating the experimental progress. Specifically, in a preferred embodiment, the above step S201, establishing the initial degradation model, includes: Randomly set the parameters in the degradation model; By individually loading physical fields onto each loading unit, and based on the real-time physical field data of the loading unit and the measured data of the mud pump life quantification index after individually loading the physical fields, the random parameters in the degradation model are corrected to obtain the initial degradation model.

[0027] The degradation model and its update process in this invention will be described in more detail below: In a preferred embodiment, the degradation model includes multiple coupled operation layers, with multiple degradation operation modules distributed in the multiple coupled operation layers. The number of coupled degradation factors represented by each coupled operation layer increases progressively. Degradation operation modules representing the same degradation factor coupling relationship are connected between adjacent coupled operation layers, and the output data of the degradation operation module of the previous layer is used as the input data of the degradation operation module of the next connected layer. The degradation operation module of the first layer is used to input its corresponding type of physical field data and mud pump state data, and the degradation operation module of the last layer is used to output prediction data.

[0028] The hierarchical, progressively coupled computational layer structure designed in this embodiment achieves a refined decomposition and orderly fusion of the degradation effects of multiple physics fields through a progressively advancing logic. Specifically, the first-layer single-factor computation module focuses on learning the independent effects of each influencing factor, the second-layer pairwise coupling modules focus on the interaction effects of two factors, and so on. This hierarchical design ensures clear boundaries for each degradation computation module while avoiding feature confusion caused by simultaneous input of multiple factors. Adjacent layers are connected in series through coupling modules of the same factor (e.g., the output of the abrasion module in the previous layer is connected to the "abrasion-aging" or "abrasion-cavitation" coupling module in the next layer), forming a data flow path from independent effects to interaction enhancement and then to comprehensive output. This preserves the basic characteristics of single factors while gradually injecting coupling information through hierarchical transmission, ensuring that the model can clearly trace each degradation index.

[0029] Specifically, combined Figure 4 As shown, based on the above model structure, the above steps: S301, correcting the degradation calculation module in the degradation model according to real-time physical field data and measured data, specifically include: S401. Based on the degradation model, obtain the computational path of the error-dominant module; S402, Freeze other degradation operation modules outside the operation path in the degradation model; S403. Based on real-time physical field data and measured data, correct the degradation calculation module in the calculation path.

[0030] In the above process, the computational path refers to the data flow path from the single-factor degradation computational module to the full-factor degradation computational module, including the error-dominant module. Once the error-dominant module is identified, its complete computational path can be traced back in reverse (from the input single-factor module, through the multi-factor coupling module, to the associated nodes in the output layer), and other irrelevant modules outside the path can be frozen. At this point, the model only needs to specifically correct the parameters within this path based on real-time physical field data and measured indicators. This avoids redundant calculations and interference from secondary factors in the full model update, and allows resources to be focused on enhancing the accuracy of understanding specific interaction relationships (such as how increased temperature and humidity exacerbate cavitation-induced fatigue cracks in blades). This path-locking mechanism combined with local updates enables the model to accurately isolate key error sources in complex coupled networks, eliminating the need for repeated full model training and endowing the model with a clear interpretive ability of physical meaning through a clear path tracing logic.

[0031] Understandably, in the initial state, the last layer, representing the degenerate computation module with comprehensive coupling of all factors, can be assumed to be the error-dominant module. At this time, the entire degenerate model is the computation path corresponding to this error-dominant module. The loading combination mode of the physical field can be dynamically changed according to the error-dominant module. To avoid the "deadlock" situation where the error-dominant module remains unchanged or loops between certain specific modules, in practice, all loading units can first load the physical field and update the degenerate model. When the error-dominant module is detected, the loading combination mode is adjusted according to the error-dominant module, and the degenerate computation modules on the corresponding path are updated. When the update reaches the preset conditions (such as the error of all degenerate computation modules on the path is lower than the set threshold, or the number of iterations reaches the set upper limit), all physical fields are loaded again and the above process is repeated.

[0032] Furthermore, in a preferred embodiment, step S403, correcting the degradation calculation module in the calculation path based on real-time physical field data and measured data, specifically includes: Real-time physical field data is input into the degradation model to obtain the predicted data output by the degradation model; The error function is obtained based on the difference between the predicted data and the measured data; Based on the error function, the degenerate computation modules in the computation path are corrected by gradient descent.

[0033] In this embodiment, the structure of the degradation model is naturally adapted to the gradient-driven error correction mechanism. When the error of a certain layer of coupled modules is significant (such as the prominent gradient of the second-layer "aging-cavitation" module), the corresponding interaction relationship can be directly located for targeted data supplementation and parameter updates without reconstructing the entire model. This significantly improves the analytical efficiency and interpretability of multi-factor coupling effects. By propagating the error back along the computation path using the gradient descent algorithm, only the parameters of the degradation computation modules within the path are updated. This achieves local optima with minimal computational cost and avoids global parameter perturbations that could destroy converged single-factor or low-order coupling knowledge. This allows the model to efficiently approximate the real degradation laws, making it particularly suitable for the fine analysis of complex interaction effects in strongly coupled multi-physics scenarios.

[0034] A more detailed embodiment is provided below to help understand the degradation model and update method in this invention.

[0035] Combination Figure 5 As shown, in a preferred embodiment, the degradation factors include abrasion, aging, and cavitation; the multiple coupled computation layers include a single-factor computation layer, a two-factor coupling layer, and a three-factor coupling layer. The degradation computation module in the single-factor computation layer is used to input the physical field data and mud pump status data of its corresponding type and output a first intermediate vector. The degradation computation module in the two-factor coupling layer is used to input the first intermediate vector corresponding to the coupled degradation factor it represents and output a second intermediate vector. The three-factor coupling layer is used to input all the second intermediate vectors and output prediction data.

[0036] More specifically, the degradation model in this embodiment can adopt a hierarchical architecture similar to a neural network. Its single-factor operation layer includes three degradation operation modules representing the effects of abrasion, aging, and cavitation, respectively, as expressed in the formula: in, , and These represent the first intermediate vectors corresponding to abrasion, aging, and cavitation, respectively. This represents the sand-bearing abrasion parameters in the physical field data. This represents the temperature, humidity, and vibration aging parameters in the physical field data. This represents the cavitation pressure fluctuation parameter in the physical field data. This indicates the basic status of the mud pump, such as its speed and cumulative running time. , and These represent the calculation methods of the degradation calculation modules corresponding to abrasion, aging, and cavitation, respectively. , and These are the parameters to be corrected for the three components. It is understandable that... , , , , and The specific format can be flexibly designed according to the actual situation. For example, a shallow neural network (such as a 1-2 layer MLP) can be used as the structure of the degenerate operation module. , , This represents matrix operations. , and This represents the parameters in the matrix (i.e., the parameters of the neural network to be trained).

[0037] The two-factor coupling layer includes three degradation calculation modules representing the effects of the two factors: erosion-aging, erosion-cavitation, and aging-cavitation, respectively. The formula is expressed as follows: in, , and These represent the second intermediate vectors corresponding to erosion-aging, erosion-cavitation, and aging-cavitation, respectively. , and These represent the calculation methods of the degradation calculation modules corresponding to abrasion-aging, abrasion-cavitation, and aging-cavitation, respectively. , and These are the parameters to be corrected for the three, and their specific calculation methods can be found in the previous explanation.

[0038] The three-factor coupling layer includes only one degradation calculation module representing the combined effects of erosion, aging, and cavitation. Its formula is expressed as: in, The output is the predicted data (in vector form). This describes the calculation method for the degradation calculation module corresponding to the abrasion-aging-cavitation process. These are the corresponding parameters. Similarly, the specific calculation method can be found in the previous explanation.

[0039] When the aforementioned different degradation operation modules are used as error-dominant modules, the corresponding operation paths (represented by the parameter symbols mentioned above) are as follows: 1. Abrasion-dominated: , , , ; 2. Aging-driven: , , , ; 3. Cavitation is the dominant factor: , , , ; 4. Abrasion-aging dominance: , , , ; 5. Abrasion-cavitation dominance: , , , ; 6. Aging-cavitation dominance: , , , ; 7. The dominant factor is the full coupling of erosion, aging, and cavitation: , , , , , , .

[0040] The error function corresponding to the above model is: in, Indicates error. This represents the total number of samples (which can be understood as the number of times data was collected in the experiment). and These are the predicted and measured data corresponding to the same physical field data, respectively. Gradient descent can be used to update the parameters of all degradation operation modules along the degradation model's computational path.

[0041] Please refer to the following: Figure 2 The present invention also provides a degradation simulation test evaluation method for the entire life cycle of mud pumps, applicable to any of the above-mentioned degradation simulation test systems for the entire life cycle of mud pumps, comprising: S201. Establish an initial degradation model. The degradation model includes multiple degradation operation modules. Each degradation operation module corresponds to one degradation factor or the degradation effect of multiple degradation factors coupled together. The degradation model is used to input mud pump status data and physical field data of each degradation factor, and outputs predicted data of mud pump life quantitative index after collaborative operation of multiple degradation operation modules. S202. After the loading unit loads the physical field, obtain the real-time physical field data of the loading unit and the measured data of the mud pump life quantification index. S203. Correct the degradation model based on real-time physical field data and measured data, and adjust the loading combination mode of multiple loading units and reload the physical field based on the error of different degradation operation modules during the degradation model correction process. S204. Repeat steps two and three to obtain the corrected degradation model, and evaluate the life of the mud pump based on the degradation model.

[0042] This invention provides a degradation simulation test system for the entire life cycle of mud pumps, comprising multiple independently controllable loading units and a test evaluation unit. Each loading unit is used to load a physical field of a degradation factor onto the mud pump. The test evaluation unit is used to establish an initial degradation model, correct the degradation model based on real-time physical field data from the loading units and measured data of the mud pump's life quantification index, and adjust the loading combination mode of multiple loading units and reload the physical field according to the error of different degradation calculation modules during the degradation model correction process. This process is repeated to obtain a corrected degradation model, and the life of the mud pump is evaluated based on the degradation model. The degradation model includes multiple degradation calculation modules, each corresponding to the degradation effect of one or more coupled degradation factors. The mud pump status data and physical field data are output as predicted data of the mud pump's life quantification index after collaborative calculation by multiple degradation calculation modules. This invention, through innovative degradation model design, decouples the independent effects of single factors and the coupled effects of multiple factors into multiple degradation calculation modules. These modules, combined with multiple independently controllable loading units, enable flexible loading of multiple factor combinations. A closed-loop mechanism, employing measured data to correct the model, model error-driven loading combination adjustment, and targeted supplementary data for further correction, locates the source of model error and dynamically switches loading unit combinations. This eliminates the interference from multiple physical fields, accurately acquires degradation data, and ultimately obtains a model that accurately characterizes the degradation laws of multi-factor coupling. This provides experimental support for the performance evolution analysis and life prediction of mud pumps throughout their entire life cycle, combining physical realism and model reliability. It solves the problem that existing mud pump test platforms cannot conduct mud pump degradation simulation tests based on multi-physical field coupling.

[0043] It should be noted that the various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. The same or similar parts between the various embodiments can be referred to each other.

[0044] The above description of the disclosed embodiments enables those skilled in the art to make or use the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A degradation simulation test system for the entire life cycle of mud pumps, characterized in that, It includes multiple independently controllable loading units and a test evaluation unit. Each loading unit is used to apply a physical field of a degradation factor to the mud pump, and the test evaluation unit is used to perform the following steps: Step 1: Establish an initial degradation model. The degradation model includes multiple degradation operation modules. Each degradation operation module corresponds to the degradation effect of one degradation factor or multiple degradation factors coupled together. The degradation model is used to input mud pump status data and physical field data of each degradation factor, and outputs predicted data of mud pump life quantitative indicators after collaborative operation of multiple degradation operation modules. Step 2: After the loading unit loads the physical field, obtain the real-time physical field data of the loading unit and the measured data of the mud pump life quantification index; Step 3: Correct the degradation model based on real-time physics data and measured data, and adjust the loading combination mode of multiple loading units and reload the physics field based on the error of different degradation operation modules during the degradation model correction process. Step 4: Repeat steps 2 and 3 to obtain the corrected degradation model, and evaluate the life of the mud pump based on the degradation model.

2. The degradation simulation test system for the entire life cycle of mud pumps according to claim 1, characterized in that, The degradation model is corrected based on real-time physics data and measured data. Furthermore, based on the errors of different degradation calculation modules during the model correction process, the loading combination mode of multiple loading units is adjusted, and the physics field is reloaded, including: The degradation calculation module in the degradation model is corrected based on real-time physical field data and measured data; Based on the correction amount of each degradation operation module, the error-dominant module is obtained; Based on the type of degradation effect corresponding to the error-dominant module, adjust the loading combination mode of multiple loading units and reload the physical field.

3. The degradation simulation test system for the entire life cycle of mud pumps according to claim 2, characterized in that, The degradation model includes multiple coupled operation layers, with multiple degradation operation modules distributed in multiple coupled operation layers. The number of coupled degradation factors represented by each coupled operation layer increases progressively. Degradation operation modules representing the same degradation factor coupling relationship are connected between adjacent coupled operation layers, and the output data of the degradation operation module of the previous layer is used as the input data of the degradation operation module of the next connected layer. The degradation operation modules of the first layer are used to input their corresponding types of physical field data and mud pump state data, and the degradation operation modules of the last layer are used to output prediction data.

4. The degradation simulation test system for the entire life cycle of mud pumps according to claim 3, characterized in that, Degradation factors include abrasion, aging, and cavitation; multiple coupled computation layers include a single-factor computation layer, a two-factor coupling layer, and a three-factor coupling layer. The degradation computation module in the single-factor computation layer is used to input the physical field data and mud pump status data of its corresponding type and output the first intermediate vector. The degradation computation module in the two-factor coupling layer is used to input the first intermediate vector corresponding to the coupled degradation factor it represents and output the second intermediate vector. The three-factor coupling layer is used to input all the second intermediate vectors and output the prediction data.

5. The degradation simulation test system for the entire life cycle of mud pumps according to claim 3, characterized in that, The degradation calculation module in the degradation model is corrected based on real-time physical field data and measured data, including: Based on the degradation model, the computational path of the error-dominant module is obtained; Freeze other degenerate operation modules outside the operation path in the degenerate model; Based on real-time physical field data and measured data, the degradation calculation module in the calculation path is corrected.

6. The degradation simulation test system for the entire life cycle of mud pumps according to claim 5, characterized in that, Based on real-time physics field data and measured data, the degenerate calculation module in the calculation path is corrected, including: Real-time physical field data is input into the degradation model to obtain the predicted data output by the degradation model; The error function is obtained based on the difference between the predicted data and the measured data; Based on the error function, the degenerate computation modules in the computation path are corrected by gradient descent.

7. The degradation simulation test system for the entire life cycle of mud pumps according to claim 2, characterized in that, Based on the correction amount of each degradation operation module, the error-dominant modules are obtained, including: The sum of the adjustments to each parameter in each degradation operation module is calculated and used as the correction amount for each degradation operation module; The degradation operation module with the highest correction value is selected as the error-dominant module.

8. The degradation simulation test system for the entire life cycle of mud pumps according to claim 1, characterized in that, Establish an initial degradation model, including: Randomly set the parameters in the degradation model; By individually loading physical fields onto each loading unit, and based on the real-time physical field data of the loading unit and the measured data of the mud pump life quantification index after individually loading the physical fields, the random parameters in the degradation model are corrected to obtain the initial degradation model.

9. The degradation simulation test system for the entire life cycle of mud pumps according to claim 1, characterized in that, Degradation factors include abrasion, aging, and cavitation; loading units include automatic sand and water supply units, temperature, humidity, and vibration integrated loading chambers, and controllable cavitation induction units; quantification indicators of mud pump life include at least one of head retention rate, shaft power fluctuation coefficient, acoustic emission spectrum characteristics, vibration amplitude change rate, and flow efficiency.

10. A degradation simulation test evaluation method for the entire life cycle of a mud pump, applied to any one of the degradation simulation test systems for the entire life cycle of a mud pump according to claims 1-9, characterized in that, include: Step 1: Establish an initial degradation model. The degradation model includes multiple degradation operation modules. Each degradation operation module corresponds to the degradation effect of one degradation factor or multiple degradation factors coupled together. The degradation model is used to input mud pump status data and physical field data of each degradation factor, and outputs predicted data of mud pump life quantitative indicators after collaborative operation of multiple degradation operation modules. Step 2: After the loading unit loads the physical field, obtain the real-time physical field data of the loading unit and the measured data of the mud pump life quantification index; Step 3: Correct the degradation model based on real-time physics data and measured data, and adjust the loading combination mode of multiple loading units and reload the physics field based on the error of different degradation operation modules during the degradation model correction process. Step 4: Repeat steps 2 and 3 to obtain the corrected degradation model, and evaluate the life of the mud pump based on the degradation model.

Images