Water scale thickness prediction method, device, equipment, medium and product

Abstract

The invention discloses a scale thickness prediction method, device and equipment, a medium and a product. The scale thickness prediction method comprises the steps that an operation data set of a cooling water system in a to-be-predicted machine room is acquired, and the operation data set comprises water hardness, total dissolved solids, the pH value, the water flow velocity, the water temperature and the operation duration; inputting the operation data set into a scale thickness prediction model to obtain a predicted water pipe scaling thickness output by the scale thickness prediction model; the incrustation thickness prediction model is obtained by taking an initial fuzzy neural network as a basic model and performing global optimization training on parameters of the basic model by adopting a Zhuang Gracula tenuifolia optimization algorithm. According to the technical scheme, the accuracy of the scaling thickness of the machine room water pipe can be improved.

Description

Technical Field

[0001] This invention relates to the field of artificial intelligence technology, and in particular to a method, apparatus, equipment, medium, and product for predicting scale thickness. Background Technology

[0002] With the rapid development of cloud computing, big data, and fintech, data centers, as the core infrastructure of modern information society, are of paramount importance in terms of operational continuity and reliability. Precision air conditioning systems in data centers are key equipment for ensuring a suitable temperature and humidity environment, and the smooth operation of their cooling water systems directly affects the overall system's heat dissipation efficiency and operational safety.

[0003] In actual operation, cooling water sources are rich in mineral ions such as calcium and magnesium. Under the combined effects of multiple factors such as temperature, flow rate, and pH value, scale easily forms on the inner walls of pipes. Scale buildup significantly reduces the thermal conductivity of pipes, increases water flow resistance, leads to increased energy consumption and decreased cooling efficiency of air conditioning units, and in severe cases, may cause localized overheating or even pipe blockage, posing a direct threat to the high availability and reliability of data centers. Therefore, accurate prediction of the thickness of scale buildup in cooling water pipes has become an important task in data center facility management. Summary of the Invention

[0004] This invention provides a method, apparatus, equipment, medium, and product for predicting scale thickness, in order to solve the problem of difficulty in modeling the nonlinear relationship between water quality parameters and scale thickness.

[0005] According to one aspect of the present invention, a method for predicting scale thickness is provided, comprising:

[0006] Obtain the operational dataset of the cooling water system in the computer room to be predicted. The operational dataset includes water hardness, total dissolved solids, pH, water flow rate, water temperature, and operating time.

[0007] The running dataset is input into the scale thickness prediction model to obtain the predicted scale thickness of the water pipes output by the scale thickness prediction model;

[0008] The scale thickness prediction model is based on an initial fuzzy neural network and is obtained by globally optimizing the parameters of the basic model using the Magnificent Slender-tailed Warbler optimization algorithm.

[0009] According to another aspect of the present invention, a scale thickness prediction device is provided, comprising:

[0010] The operation data acquisition module is used to acquire the operation data set of the cooling water system in the computer room to be predicted. The operation data set includes water hardness, total dissolved solids, pH, water flow rate, water temperature and operating time.

[0011] The scale thickness prediction module is used to input the running dataset into the scale thickness prediction model to obtain the predicted scale thickness of the water pipe output by the scale thickness prediction model.

[0012] The scale thickness prediction model is based on an initial fuzzy neural network and is obtained by globally optimizing the parameters of the basic model using the Magnificent Slender-tailed Warbler optimization algorithm.

[0013] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0014] At least one processor; and

[0015] A memory communicatively connected to the at least one processor; wherein,

[0016] The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the scale thickness prediction method according to any embodiment of the present invention.

[0017] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the scale thickness prediction method according to any embodiment of the present invention.

[0018] According to another aspect of the present invention, a computer program product is provided, comprising a computer program that, when executed by a processor, implements the scale thickness prediction method of any embodiment of the present disclosure.

[0019] The technical solution of this invention involves obtaining an operational dataset of the cooling water system in the computer room to be predicted. The operational dataset includes water hardness, total dissolved solids, pH, water flow velocity, water temperature, and operating time. The operational dataset is then input into a scale thickness prediction model to obtain the predicted scale thickness of the water pipes output by the scale thickness prediction model. The scale thickness prediction model is based on an initial fuzzy neural network and is obtained by globally optimizing and training the parameters of the basic model using the Magnificent Finch Warbler optimization algorithm. This solves the problem of difficulty in modeling the nonlinear relationship between water quality parameters and scale thickness, and improves the accuracy of predicting the scale thickness of water pipes in the computer room.

[0020] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a flowchart of a method for predicting scale thickness according to Embodiment 1 of the present invention;

[0023] Figure 2a This is a flowchart of a method for predicting scale thickness according to Embodiment 2 of the present invention;

[0024] Figure 2b This is a flowchart of the parameter optimization method provided in Embodiment 2 of the present invention;

[0025] Figure 3 This is a schematic diagram of a scale thickness prediction device provided in Embodiment 3 of the present invention;

[0026] Figure 4 This is a schematic diagram of the structure of an electronic device that implements the scale thickness prediction method of this invention. Detailed Implementation

[0027] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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 should fall within the scope of protection of the present invention.

[0028] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0029] Example 1

[0030] Figure 1This is a flowchart illustrating a method for predicting scale thickness according to Embodiment 1 of the present invention. This embodiment is applicable to situations where scale thickness prediction is performed by combining fuzzy neural networks and the Magnificent Wren algorithm. This method can be executed by a scale thickness prediction device, which can be implemented in hardware and / or software and can be configured in various general-purpose computing devices. For example... Figure 1 As shown, the method includes:

[0031] S110. Obtain the operating dataset of the cooling water system in the computer room to be predicted. The operating dataset includes water hardness, total dissolved solids, pH, water flow rate, water temperature and operating time.

[0032] The runtime dataset is a collection of multi-dimensional parameters collected or aggregated from the cooling water system of the prediction room during the prediction period, which characterize the system state and are used for scale thickness prediction. The runtime dataset includes water quality data reflecting the chemical properties of the water in the cooling water system and the system's runtime (the duration from the last pipe cleaning time to the prediction time). For example, water quality data may include water hardness, total dissolved solids, pH, flow rate, and water temperature.

[0033] In this embodiment of the invention, the operational dataset of the cooling water system in the computer room to be predicted is first obtained. Specifically, sensors installed at corresponding nodes in the cooling water pipeline collect raw water quality data at set time intervals. The collected raw water quality data includes water hardness, total dissolved solids, pH, water flow velocity, and water temperature. For each data sequence of collected water quality data, the mean value of each data type is calculated and stored in the operational dataset. Simultaneously, the continuous operating time of the cooling water system between the last pipeline cleaning time and the prediction time is determined and stored in the operational dataset. The final operational dataset includes water hardness, total dissolved solids, pH, water flow velocity, water temperature, and operating time.

[0034] S120. Input the running dataset into the scale thickness prediction model to obtain the predicted scale thickness of the water pipes output by the scale thickness prediction model.

[0035] Among them, the scale thickness prediction model is based on an initial fuzzy neural network model, and the parameters of the basic model are globally optimized and trained using the Magnificent Slender-tailed Warbler optimization algorithm.

[0036] The scale thickness prediction model is a computational model used to predict the thickness of water pipe structures at a future set time based on multi-dimensional operating data of the cooling water system. The core of the subsequent scale calculation model is a fuzzy neural network, comprising an input layer, a fuzzification layer, a rule layer, a normalization layer, and an output layer. The network's internal parameters are determined through a global optimization search using the Magnificent Wren optimization algorithm. For example, when the membership function is determined to be a Gaussian membership function, the network's internal parameters may include the center of the membership function, the width of the membership function, and the rule weights.

[0037] The Magnificent Wren Optimization Algorithm is an intelligent optimization algorithm that mimics the behavioral patterns of magnificent wren flocks at different life stages such as growth, reproduction, and defense. By simulating three strategies—"chick growth" (extensive exploration), "breeding and feeding" (local fine development), and "avoiding predators" (introducing random mutations to escape local optima)—it dynamically adjusts the search process to efficiently find the global optimum or high-performance solution in a complex high-dimensional parameter space.

[0038] In this embodiment of the invention, the obtained operating dataset of the cooling water system is input into a scale thickness prediction model, which then predicts the scale thickness in the water pipes. Specifically, the scale thickness prediction model performs the following calculation steps: After receiving the operating dataset at the input layer, it is transmitted to the fuzzification layer, which fuzzifies each data item in the operating dataset, mapping the input space to a Gaussian distribution space to obtain the membership degree of each data item to each fuzziness level. Further, the rule layer calculates the activation strength (i.e., applicability) of each pre-set rule based on the membership degree of each data item to each fuzziness level, and the normalization layer normalizes the activation strength of each rule. Finally, the output layer calculates the predicted scale thickness in the water pipes based on the normalized data of each rule. By using a fuzzy neural network globally optimized by the Magnificent Wren optimization algorithm, the model can deeply explore the complex nonlinear and fuzzy relationship between operating data and scale thickness, avoiding the shortcomings of traditional machine learning models that are prone to getting trapped in local optima and have limited prediction accuracy, thereby outputting highly reliable scale thickness prediction values.

[0039] The technical solution of this invention involves obtaining an operational dataset of the cooling water system in the computer room to be predicted. The operational dataset includes water hardness, total dissolved solids, pH, water flow velocity, water temperature, and operating time. The operational dataset is then input into a scale thickness prediction model to obtain the predicted scale thickness of the water pipes output by the scale thickness prediction model. The scale thickness prediction model is based on an initial fuzzy neural network and is obtained by globally optimizing and training the parameters of the basic model using the Magnificent Finch Warbler optimization algorithm. This solves the problem of difficulty in modeling the nonlinear relationship between water quality parameters and scale thickness, and improves the accuracy of predicting the scale thickness of water pipes in the computer room.

[0040] Example 2

[0041] Figure 2a This is a flowchart of a scale thickness prediction method provided in Embodiment 2 of the present invention. This embodiment further refines the above embodiment, providing specific steps for inputting the running dataset into the scale thickness prediction model to obtain the predicted scale thickness of the water pipes output by the model, and specific steps for globally optimizing the parameters of the basic model using the Magnificent Slender-tailed Warbler optimization algorithm. Figure 2a As shown, the method includes:

[0042] S210. Obtain the operating dataset of the cooling water system in the computer room to be predicted. The operating dataset includes water hardness, total dissolved solids, pH, water flow rate, water temperature and operating time.

[0043] S220. Input the running dataset into the scale thickness prediction model. The input layer transmits the running dataset to the fuzzing layer, and the fuzzing layer performs fuzzing processing on each data item in the running dataset to obtain the membership degree of each data item to each fuzzing level.

[0044] The fuzzification layer contains a series of predefined membership functions. The fuzziness level is a qualitative description of the state of an input variable. For example, for the input variable "water temperature," three fuzziness levels can be set: "low temperature," "medium temperature," and "high temperature," each corresponding to a temperature range. The membership degree is a value between 0 and 1, used to quantify the degree to which the precise value of an input variable belongs to a specific fuzziness level. For example, a water temperature of 32°C has a membership degree of 0.8 for the fuzziness level "high temperature."

[0045] In this embodiment of the invention, the running dataset is first input into the scale thickness prediction model. The running dataset X is as follows:

[0046] ;

[0047] in, It's about water hardness. It is the total dissolved solids. It's about pH level. It is the water flow velocity. It's the water temperature. This is the runtime; the number of nodes in the input layer is 6.

[0048] The input layer of the scale thickness prediction model transmits the running dataset to the fuzzification layer. The fuzzification layer pre-defines several fuzziness levels for each data item in the running dataset. For example, for the input variable "water temperature," three fuzziness levels can be set: "low temperature," "medium temperature," and "high temperature." Each fuzziness level corresponds to a membership function. , The membership function can be specifically expressed as:

[0049] ;

[0050] in, It represents the membership degree of the i-th data item in the dataset at the j-th fuzzy level it is associated with. It is the width of the membership function. t is the center value of the membership function, and t is the number of iterations of the optimization algorithm for the magnificent slender-tailed warbler.

[0051] The fuzzing layer can take each data item in the input running dataset and input it into the membership function corresponding to all fuzziness levels of that data item to obtain the membership degree of each data item to each fuzziness level.

[0052] In a specific example, the fuzzification layer's "water temperature" variable is preset with three fuzziness levels: "low," "medium," and "high." When the input water temperature is 32℃, the fuzzification layer will calculate the membership degree of the input water temperature to the three fuzziness levels in parallel. For example, the calculation results are 0.1, 0.4, and 0.8, respectively. That is, in the current model, a water temperature of 32℃ is quantified as belonging to "low temperature" with a degree of 0.1, to "medium temperature" with a degree of 0.4, and to "high temperature" with a degree of 0.8.

[0053] Optionally, the fuzzing layer uses a Gaussian membership function to fuzzify each data item in the running dataset. The shape of the Gaussian membership function is determined by its center value and width.

[0054] In this embodiment of the invention, the membership function used in the fuzzification layer is defined as a Gaussian membership function. The important parameters in the Gaussian membership function are the center value of the membership function and the width of the membership function.

[0055] S230. The rule layer determines the applicability of each fuzzy rule based on the membership degree of each data item to each fuzzy level, and the normalization layer performs a normalization operation on the applicability of each fuzzy rule to obtain the normalized data of each fuzzy rule.

[0056] The rule layer consists of multiple nodes, each node corresponding to a fuzzy rule in the form of "if-then", where the "if" part is the premise and the "then" part is the reasoning conclusion.

[0057] In this embodiment of the invention, the rule layer receives the membership degrees from the fuzzification layer, and then extracts the membership degrees associated with all preconditions for each rule. Based on the extracted membership degrees, the applicability of the rule is calculated. The specific calculation formula for the k-th fuzzy rule is as follows:

[0058] ;

[0059] in, It is the output value of the kth node in the rule layer, that is, the applicability of the kth fuzzy rule.

[0060] Furthermore, the applicability of each fuzzy rule is normalized by the normalization layer to obtain the normalized data for each fuzzy rule. The specific normalization formula is as follows:

[0061] ;

[0062] in, It is the output value of the kth node of the normalization layer, that is, the normalized data of the kth fuzzy rule.

[0063] S240. From the output layer, based on the output weight of each fuzzy rule, the normalized data of multiple fuzzy rules are weighted and summed to obtain the predicted thickness of scale buildup in the water pipe.

[0064] In this embodiment of the invention, the output layer performs a weighted summation of the normalized data of multiple fuzzy rules based on the output weights corresponding to each fuzzy rule to obtain the predicted thickness of scale buildup in the water pipe. The specific calculation formula is as follows:

[0065] ;

[0066] in, It is the output weight corresponding to the k-th fuzzy rule. It is the output value of the kth node in the normalization layer.

[0067] The fuzzification layer of the scale thickness prediction model establishes an independent fuzzy description for each variable, decoupling the complex relationships between variables. Then, the rule layer establishes an interaction network between variables through the combination of rules. Finally, the normalization and output layer balances and integrates all the outputs of the network, enabling the model to learn the nonlinear physicochemical mechanism of the scaling process, giving full play to the nonlinear fitting ability of the fuzzy neural network and improving the prediction accuracy.

[0068] Optionally, the scale thickness prediction model is based on an initial fuzzy neural network model, and the parameters of the basic model are globally optimized and trained using the Magnificent Wren optimization algorithm. The global optimization of the parameters of the basic model using the Magnificent Wren optimization algorithm includes:

[0069] A set of candidate solutions is randomly generated, and each candidate solution is a combination of parameters of the basic model;

[0070] In each iteration, a random number is generated for each candidate solution, and a risk threshold for each candidate solution is determined based on the normally distributed random number. The growth stage of each candidate solution is determined based on the random number and risk threshold. The growth stages include the chick growth stage, the breeding and raising stage, and the predator avoidance stage.

[0071] The candidate solutions are updated according to the position update strategy corresponding to the growth stage of each candidate solution. The updated candidate solutions are used as the model parameters of the basic model. The scale thickness is predicted based on the historical transportation dataset, and the error between the predicted scale thickness and the actual scale thickness is determined. The individual optimal solution and the global optimal solution are updated based on the error.

[0072] When the preset maximum number of iterations is reached, the parameter combination corresponding to the global optimal solution is determined as the optimal parameters of the fuzzy neural network model.

[0073] In this optional embodiment, a specific method is provided for globally optimizing the parameters of the basic model using the Magnificent Slender-tailed Warbler optimization algorithm, such as... Figure 2b As shown: First, initialize the population. Within the preset parameter value range, randomly generate a group of candidate solutions. Each candidate solution in the group is a high-dimensional vector, whose dimension is equal to the total number of all parameters to be optimized in the fuzzy neural network, which is a parameter combination of the basic model.

[0074] Then, multiple rounds of iterative updates are performed on each candidate solution. In each iteration, a random number is generated for each candidate solution, and a risk threshold is determined for each candidate solution based on the normally distributed random number. Then, based on the random number and risk threshold of each candidate solution, the growth stage of each candidate solution is determined. The growth stages include the chick growth stage, the breeding and rearing stage, and the predator avoidance stage.

[0075] The position update strategy for the chick growth stage tends to superimpose a large-scale random perturbation around the current solution, prompting candidate solutions to explore previously unexplored areas in the parameter space. The position update strategy for the breeding and rearing stage tends to make small, precise movements towards the currently known global optimum or its own individual optimum, aiming to deeply explore high-quality areas. The position update strategy for the predator avoidance stage, based on learning from superior individuals, introduces a large, random step size following a heavy-tailed distribution, such as the Lévy flight pattern, thus allowing candidate solutions to potentially escape their current local optima and seek new, more optimal areas.

[0076] Furthermore, a corresponding position update strategy is determined based on the growth stage of each candidate solution, and the corresponding candidate solutions are updated accordingly. The updated candidate solutions are used as the model parameters of the base model. The scale thickness is predicted based on historical transportation datasets, and the error between the predicted scale thickness and the actual scale thickness is determined. The individual optimal solution and the global optimal solution are updated based on the error. Finally, when the number of iterations reaches the preset maximum number of iterations, the parameter combination corresponding to the global optimal solution is determined as the optimal parameter of the fuzzy neural network model.

[0077] The Magnificent Wren-Wren optimization algorithm replaces the cumbersome, inefficient, and easily localized parameter determination methods that rely on expert experience or grid search. By simulating the intelligent behavior of biological groups, it can efficiently explore in high-dimensional and complex parameter spaces, improve the model parameter optimization performance, and the mutation strategy introduced in the predator avoidance stage can escape local optima, thereby ensuring the most effective and stable model performance.

[0078] Optionally, if the candidate solution is in the predator avoidance stage of its growth, the candidate solution is updated according to the position update strategy corresponding to the growth stage, including:

[0079] Generate a Lévy flight stochastic step size that conforms to the Lévy distribution, and determine an adaptive balance factor based on the ratio between the current iteration number and the maximum iteration number;

[0080] The candidate solution is updated based on the current global optimal solution, the current position of the candidate solution, the Lévy flight stochastic step size, and the adaptive balance factor for iterative process management.

[0081] In this optional embodiment, a specific method is provided for updating candidate solutions according to the position update strategy corresponding to the growth stage when the candidate solution is in the predator avoidance stage: A Lévy flight stochastic step size conforming to the Lévy distribution is generated, and an adaptive balance factor is determined based on the ratio between the current iteration number and the maximum iteration number. The candidate solution is then updated based on the current global optimal solution, the current position of the candidate solution, the Lévy flight stochastic step size, and the adaptive balance factor managed by the iteration process. Specifically, during the predator avoidance stage, the position update of individuals in the population is based on the defense mechanism of the Magnificent Swan-warbler against predator attacks. When a Magnificent Swan-warbler is spotted by a predator, it will run quickly, flapping its wings to disrupt the predator's line of sight and emitting alarm calls to alert other Magnificent Swan-warbler individuals. In this situation, the Magnificent Swan-warbler targeted by the predator will quickly escape, causing a slight change in its position; while other individuals will hover in the air to avoid the predator, resulting in a significant change in their position. These two different behavioral modes enhance the algorithm's search range in the problem-solving space and its ability to utilize local searches. The description formula for their movement is as follows:

[0082] ;

[0083] in, It represents the position of the i-th magnificent swan-warbler in the j-th dimension after the t-th iteration. It is a Lévy flight stochastic step size that conforms to the Lévy distribution, used to control the algorithm to escape local optima. It is an adaptive balance factor, and Adjusting flight distance together It represents the current globally optimal solution and is used to control the movement direction of individuals, preventing them from moving towards undesirable positions. It is a random number within the interval [0,1]. It is a risk threshold.

[0084] The adaptive balance factor is calculated as follows:

[0085] ;

[0086] Among them, the w-dimensional call frequency value serves as an early warning during flight to avoid predators.

[0087] The risk threshold is calculated as follows:

[0088] ;

[0089] in, and All are random numbers that follow a normal distribution.

[0090] By introducing a position update strategy corresponding to the predator avoidance phase, and based on learning from excellent individuals, the stochastic step size of Levy's flight is introduced, which makes it possible for candidate solutions to jump out of the current local optimum and find new potential better regions.

[0091] Optionally, if the candidate solution is in the breeding and rearing stage of its growth, the candidate solution is updated according to the position update strategy corresponding to the growth stage, including:

[0092] Based on the current global optimal solution, the maturity factor that changes periodically with the iteration process, and the current position of the candidate solution, the candidate solution is updated.

[0093] In this optional embodiment, a specific method is provided for updating candidate solutions according to the position update strategy corresponding to the growth stage when the candidate solution is in the breeding and rearing stage: The candidate solution is updated based on the current global optimal solution, the maturity factor that changes periodically with the iteration process, and the current position of the candidate solution. Specifically, in the breeding and rearing stage, the position of individuals in the population is updated by simulating the "teaching mechanism" of the Magnificent Swan-warbler during breeding and chick rearing. When the risk threshold is low, the Magnificent Swan-warbler will enter the breeding stage and employ a unique parentage identification mechanism during incubation to prevent the invasion of alien species.

[0094] Because the Magnificent Swan-tailed Warbler exhibits cooperative breeding, multiple individuals incubate the eggs throughout the year to teach offspring recognition. During this teaching cycle (m), each individual Swan-tailed Warbler is not stationary but takes turns feeding and teaching. Modeling this phenomenon leads to slight changes in the individual's position, thus enhancing the algorithm's ability to explore local areas. Simultaneously, a maturity factor p is defined, which gradually increases as the teaching cycle shortens: the closer to maturity, the larger the activity range of each individual Swan-tailed Warbler. Based on modeling the positional changes during the teaching and incubation period, if the new position of each individual improves the objective function value, the corresponding individual's current position will be updated. The specific formula for position updating is as follows:

[0095] ;

[0096] ;

[0097] in, It represents the position of the i-th magnificent swan-warbler in the j-th dimension after the t-th iteration. This is the current globally optimal solution, where p is the maturity factor and C is a constant. It is a random number within the interval [0,1]. It is a risk threshold.

[0098] The formula for calculating the maturity factor is as follows:

[0099] ;

[0100] ;

[0101] Where ub and lb are the upper and lower bounds of each dimension in the running dataset, respectively, FEs is the current number of iterations, and MaxFEs is the maximum number of iterations.

[0102] The position update strategy corresponding to the breeding and rearing stage tends to make candidate solutions move slightly and finely toward the currently known global optimal solution or their own individual optimal solution, which can deeply explore high-quality areas.

[0103] In addition, during the chick development stage of the Magnificent Swan-tailed Warbler, the position updates of individuals within the population are based on a dynamic simulation of the extensive experience required for chick growth. Since a large number of chicks in the population is detrimental to the survival of the entire population, the position of individual Magnificent Swan-tailed Warblers in the problem space is simulated by continuously updating their positions as chicks accumulate experience during rapid growth. This learning process from experience is modeled as a series of positional changes, designed to induce widespread changes in the positions of individual Magnificent Swan-tailed Warblers, thereby enhancing the exploration capability of the global search algorithm. The position update formula for the chick development stage is as follows:

[0104] ;

[0105] in, is the position of the i-th magnificent swan warbler in the j-th dimension after the t-th iteration, ub and lb are the upper and lower limits of each dimension in the running dataset, respectively, and rand is a random number in the interval [0,1].

[0106] The position update strategy corresponding to the chick growth stage tends to superimpose a large-scale random perturbation near the current solution, prompting candidate solutions to explore regions in the parameter space that have not been explored before.

[0107] S250. Compare the predicted scale thickness in the water pipe with the preset safety threshold.

[0108] In this embodiment of the invention, the predicted scale thickness of the water pipe is compared with a preset safety threshold to determine whether the current scale thickness will affect the cooling water system of the computer room to be predicted.

[0109] S260. When the thickness of scale buildup in water pipes is predicted to exceed the safety threshold, an early warning message is generated.

[0110] In this embodiment of the invention, an early warning message is generated when the predicted scale thickness in the water pipe exceeds a safe threshold. The early warning message may include the predicted scale thickness and the predicted time point, reminding staff to take descaling measures before the predicted time point to avoid excessive scale affecting the operation of the cooling water system, thereby ensuring the reliability of the computer room operation.

[0111] The technical solution of this invention involves obtaining an operational dataset of the cooling water system in the computer room to be predicted. The operational dataset includes water hardness, total dissolved solids, pH, water flow velocity, water temperature, and operating time. The operational dataset is then input into a scale thickness prediction model to obtain the predicted scale thickness of the water pipes output by the scale thickness prediction model. The scale thickness prediction model is based on an initial fuzzy neural network and is obtained by globally optimizing and training the parameters of the basic model using the Magnificent Finch Warbler optimization algorithm. This solves the problem of difficulty in modeling the nonlinear relationship between water quality parameters and scale thickness, and improves the accuracy of predicting the scale thickness of water pipes in the computer room.

[0112] Example 3

[0113] Figure 3 This is a schematic diagram of a scale thickness prediction device provided in Embodiment 3 of the present invention. Figure 3 As shown, the device includes:

[0114] The operation data acquisition module 310 is used to acquire the operation data set of the cooling water system in the computer room to be predicted. The operation data set includes water hardness, total dissolved solids, pH, water flow rate, water temperature and operating time.

[0115] The scale thickness prediction module 320 is used to input the running dataset into the scale thickness prediction model to obtain the predicted scale thickness of the water pipe output by the scale thickness prediction model.

[0116] The scale thickness prediction model is based on an initial fuzzy neural network and is obtained by globally optimizing the parameters of the basic model using the Magnificent Slender-tailed Warbler optimization algorithm.

[0117] The technical solution of this invention involves obtaining an operational dataset of the cooling water system in the computer room to be predicted. The operational dataset includes water hardness, total dissolved solids, pH, water flow velocity, water temperature, and operating time. The operational dataset is then input into a scale thickness prediction model to obtain the predicted scale thickness of the water pipes output by the scale thickness prediction model. The scale thickness prediction model is based on an initial fuzzy neural network and is obtained by globally optimizing and training the parameters of the basic model using the Magnificent Finch Warbler optimization algorithm. This solves the problem of difficulty in modeling the nonlinear relationship between water quality parameters and scale thickness, and improves the accuracy of predicting the scale thickness of water pipes in the computer room.

[0118] Optionally, the scale thickness prediction model includes an input layer, a fuzzification layer, a rule layer, a normalization layer, and an output layer;

[0119] Scale thickness prediction module 320, specifically used for:

[0120] The running dataset is input into the scale thickness prediction model. The input layer transmits the running dataset to the fuzzing layer, and the fuzzing layer performs fuzzing processing on each data item in the running dataset to obtain the membership degree of each data item to each fuzzing level.

[0121] The rule layer determines the applicability of each fuzzy rule based on the membership degree of each data item to each fuzzy level, and the normalization layer normalizes the applicability of each fuzzy rule to obtain the normalized data of each fuzzy rule.

[0122] The output layer performs a weighted summation of the normalized data of multiple fuzzy rules based on the output weight of each fuzzy rule to obtain the predicted thickness of scale buildup in the water pipe.

[0123] Optionally, the scale thickness prediction module further includes a parameter optimization module, wherein the parameter optimization module includes:

[0124] The candidate solution population generation unit is used to randomly generate a group of candidate solutions, where each candidate solution is a parameter combination of the basic model.

[0125] The growth stage determination unit is used to generate random numbers for each candidate solution in each iteration, determine the risk threshold of each candidate solution based on the normally distributed random numbers, and determine the growth stage of each candidate solution based on the random numbers and risk thresholds of each candidate solution; the growth stages include the chick growth stage, the breeding and raising stage, and the predator avoidance stage;

[0126] The candidate solution update unit is used to update the candidate solution according to the position update strategy corresponding to the growth stage of each candidate solution. The updated candidate solution is used as the model parameter of the basic model. The scale thickness is predicted according to the historical transportation dataset, and the error between the predicted scale thickness and the actual scale thickness is determined. The individual optimal solution and the global optimal solution are updated according to the error.

[0127] The optimal parameter determination unit is used to determine the parameter combination corresponding to the global optimal solution as the optimal parameters of the fuzzy neural network model when the preset maximum number of iterations is reached.

[0128] Optionally, when the candidate solution is in the predator avoidance stage of its growth, the candidate solution update unit is specifically used for:

[0129] Generate a Lévy flight stochastic step size that conforms to the Lévy distribution, and determine an adaptive balance factor based on the ratio between the current iteration number and the maximum iteration number;

[0130] The candidate solution is updated based on the current global optimal solution, the current position of the candidate solution, the Lévy flight stochastic step size, and the adaptive balance factor for iterative process management.

[0131] Optionally, when the candidate solution is in the breeding and rearing stage of its growth, the candidate solution update unit is specifically used for:

[0132] Based on the current global optimal solution, the maturity factor that changes periodically with the iteration process, and the current position of the candidate solution, the candidate solution is updated.

[0133] Optionally, the fuzzification layer uses a Gaussian membership function to fuzzify each data item in the running dataset, and the shape of the Gaussian membership function is determined by its center value and width.

[0134] Optional, the scale thickness prediction device also includes:

[0135] The scale thickness comparison module is used to compare the predicted scale thickness of the water pipe with a preset safety threshold after inputting the running dataset into the scale thickness prediction model and obtaining the predicted scale thickness of the water pipe output by the scale thickness prediction model.

[0136] The early warning module is used to generate early warning information when the predicted scale thickness in the water pipe is greater than the safety threshold.

[0137] The scale thickness prediction device provided in this embodiment of the invention can execute the scale thickness prediction method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0138] In the technical solution of this invention, the information collected is information and data authorized by the user or fully authorized by all parties. The collection, storage, use, processing, transmission, provision, disclosure and application of related data all comply with the relevant laws, regulations and standards of relevant countries and regions, take necessary confidentiality measures, do not violate public order and good morals, and provide corresponding operation entry points for users to choose to authorize or refuse.

[0139] Example 4

[0140] According to embodiments of the present invention, the present invention also provides an electronic device, a readable storage medium, and a computer program product.

[0141] Figure 4A schematic diagram of an electronic device 10 that can be used to implement embodiments of the present invention is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, application processors, blade application processors, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0142] like Figure 4 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory 12 or a random access memory 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the read-only memory 12 or a computer program loaded from storage unit 18 into the random access memory 13. The random access memory 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, read-only memory 12, and random access memory 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0143] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0144] Processor 11 can be a variety of general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, central processing units, graphics processing units, various special-purpose artificial intelligence computing chips, various processors running machine learning model algorithms, digital signal processors, and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as the scale thickness prediction method.

[0145] In some embodiments, the scale thickness prediction method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via read-only memory 12 and / or communication unit 19. When the computer program is loaded into random access memory 13 and executed by processor 11, one or more steps of the scale thickness prediction method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the scale thickness prediction method by any other suitable means (e.g., by means of firmware).

[0146] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays, application-specific integrated circuits (ASICs), application-specific standard products (ASICs), systems-on-a-chip (SoCs), complex programmable logic devices, computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0147] Computer programs used to implement the methods of the present invention can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs can be executed entirely on a machine, partially on a machine, as a standalone software package partially on a machine and partially on a remote machine, or entirely on a remote machine or application.

[0148] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory, read-only memory, erasable programmable read-only memory, optical fibers, portable compact disk read-only memory, optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0149] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a cathode ray tube or liquid crystal display monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0150] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data application processors), or computing systems that include middleware components (e.g., application application processors), or computing systems that include frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0151] A computing system can include clients and applications. Clients and applications are generally geographically separated and typically interact via communication networks. The client-application relationship is established by computer programs running on the respective computers and having a client-application relationship with each other. An application can be a cloud application, also known as a cloud computing application or cloud host, which is a host product within the cloud computing application architecture to address the shortcomings of traditional physical hosts and virtual private services, such as high management difficulty and weak business scalability.

[0152] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0153] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A scale thickness prediction method characterized by, The method comprises the following steps: acquiring an operation data set of a cooling water system in a machine room to be predicted, the operation data set comprising water hardness, total dissolved solids, pH, water flow rate, water temperature, and operation duration; inputting the operation data set into a scale thickness prediction model to obtain a predicted scale thickness of the water pipe output by the scale thickness prediction model; wherein the scale thickness prediction model is based on an initial fuzzy neural network as a basic model, and a global optimization training is performed on parameters of the basic model by using a splendid sylviella optimization algorithm.

2. The method of claim 1, wherein, The scale thickness prediction model comprises an input layer, a fuzzification layer, a rule layer, a normalization layer, and an output layer. Inputting the operation data set into the scale thickness prediction model to obtain a predicted scale thickness of the water pipe output by the scale thickness prediction model comprises: inputting the operation data set into the scale thickness prediction model, transmitting the operation data set to the fuzzification layer by the input layer, and performing fuzzification processing on each item of data in the operation data set by the fuzzification layer to obtain the membership of each item of data to each fuzzy level; determining the applicability of each fuzzy rule based on the membership of each item of data to each fuzzy level by the rule layer, and performing normalization operation on the applicability of each fuzzy rule by the normalization layer to obtain the normalized data of each fuzzy rule; performing weighted summation on the normalized data of multiple fuzzy rules based on the output weight of each fuzzy rule by the output layer to obtain the predicted scale thickness of the water pipe.

3. The method of claim 1, wherein, Performing global optimization on the parameters of the basic model by using the splendid sylviella optimization algorithm comprises: randomly generating a group of candidate solution populations, each candidate solution being a parameter combination of the basic model; in each iteration, generating a random number for each candidate solution, and determining a risk threshold of each candidate solution based on the normal distribution random number, determining the growth stage of each candidate solution according to the random number and the risk threshold of each candidate solution; the growth stage comprises the chick growth stage, the breeding and feeding stage, and the predator avoidance stage; updating the candidate solution according to the position update strategy corresponding to the growth stage of each candidate solution, using the updated candidate solution as the model parameter of the basic model, predicting the scale thickness according to the historical transportation data set, and determining the error between the predicted scale thickness and the actual scale thickness, updating the individual optimal solution and the global optimal solution according to the error; when the preset maximum number of iterations is reached, determining the parameter combination corresponding to the global optimal solution as the optimal parameter of the fuzzy neural network model.

4. The method of claim 3, wherein, In the case that the growth stage of the candidate solution is the predator avoidance stage, the candidate solution is updated according to the position update strategy corresponding to the growth stage of the candidate solution, comprising: generating a Levy flight random step conforming to the Levy distribution, and determining an adaptive balance factor based on the proportional relationship between the current number of iterations and the maximum number of iterations; updating the candidate solution based on the current global optimal solution, the current position of the candidate solution, the Levy flight random step, and the adaptive balance factor of the iteration progress management.

5. The method of claim 3, wherein, In the case that the growth stage of the candidate solution is the propagation feeding stage, the candidate solution is updated according to the position updating strategy corresponding to the growth stage of the candidate solution, including: Based on the current global optimal solution, the mature factor periodically changed with the iteration process, and the current position of the candidate solution, the candidate solution is updated.

6. The method of claim 2, wherein, The fuzzy layer adopts a Gaussian membership function to perform fuzzy processing on each item of data in the operation data set, and the shape of the Gaussian membership function is determined by a center value and a width.

7. The method of claim 1, wherein, After inputting the operation data set into the scale thickness prediction model and obtaining the predicted water pipe scale thickness output by the scale thickness prediction model, the method further includes: Comparing the predicted water pipe scale thickness with a preset safety threshold value; In the case that the predicted water pipe scale thickness is greater than the safety threshold value, generating a warning information.

8. A scale thickness prediction device characterized by comprising: The method comprises: an operation data acquisition module configured to acquire an operation data set of a cooling water system in a machine room to be predicted, the operation data set comprising water hardness, total dissolved solids, pH, water flow rate, water temperature, and operation duration; a scale thickness prediction module configured to input the operation data set into a scale thickness prediction model to obtain a predicted water pipe scale thickness output by the scale thickness prediction model; wherein the scale thickness prediction model is based on an initial fuzzy neural network as a base model, and a beautiful fine tail wren optimization algorithm is used to globally optimize and train parameters of the base model.

9. An electronic device, comprising: The electronic device comprises: at least one processor; and a memory connected in communication with the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to execute the scale thickness prediction method of any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores computer instructions for causing the processor to execute the scale thickness prediction method of any one of claims 1-7 when executed.

11. A computer program product, characterised in that, The computer program product comprises a computer program, and the computer program implements the scale thickness prediction method according to any one of claims 1-7 when executed by the processor. The computer program product comprises a computer program, and the computer program implements the scale thickness prediction method according to any one of claims 1-7 when executed by the processor.

Images