A cooling tower all-around safety protection and failure early warning system

Abstract

The present application relates to the technical field of industrial management, in particular to a cooling tower all-around safety protection and fault early warning system, comprising a "sound-chemical" data acquisition module, which acquires "sound-chemical" data sets; a corrosion risk identification module, which establishes a "sound-chemical" coupling model, carries out corrosion risk assessment on a target area, and outputs spatial positioning information; a responsibility node matching module, which automatically matches responsibility nodes and constructs an initial correlation between "corrosion points and responsibility nodes"; a cross responsibility identification and optimal response path generation module, which calculates the responsibility coverage conflict degree of each responsibility node by using a responsibility coverage conflict degree measurement model, and judges whether there is a cross node; if there is a cross node, a comprehensive scheduling scoring model is constructed, and based on the task response time delay, the load balancing index and the node priority weight, an optimal response path is generated; a task scheduling module, which according to the optimal response path, dispatches maintenance tasks to corresponding responsibility nodes in priority order for hierarchical collaborative response management.

Description

Technical Field

[0001] This invention relates to the field of industrial management technology, specifically to a comprehensive safety protection and fault early warning system for cooling towers. Background Technology

[0002] In large-scale industrial production scenarios, cooling towers are typically distributed across multiple production stages. As one of the energy-intensive facilities, their operational stability has an indirect but significant impact on the overall operational efficiency of the enterprise. Due to long operating hours and high environmental exposure, cooling towers and their auxiliary structures often suffer from corrosion, aging, and material fatigue, requiring periodic monitoring and maintenance management across departments.

[0003] Currently, enterprises typically use local monitoring equipment (such as noise, vibration, and water quality recorders) to collect cooling tower operation data. However, this data is mostly stored in a "dispersed and passively analyzed" state at the system level, making it difficult to effectively integrate into the enterprise's industrial data management platform and also unable to efficiently support cross-departmental task decision-making and response coordination.

[0004] Furthermore, in actual management, addressing the corrosion risks of cooling towers often involves multiple responsible parties, including operations, equipment, water treatment, and outsourced maintenance. Due to the lack of a complete responsibility identification, conflict resolution, and task allocation mechanism, the following management bottlenecks are easily created: Unclear overlapping responsibilities: Multiple departments may respond repeatedly to the same potential hazard area or pass the buck, leading to task conflicts or a vacuum in handling the issue; Rigid scheduling mechanisms: The lack of dynamic optimization paths based on risk level, resource availability, and task relevance easily leads to resource misallocation and response delays; Low data utilization efficiency: Acoustic, water quality, and other data are only used for equipment status assessment and fail to rise to the organizational level as a basis for task triggering and responsibility decision-making, making it difficult to achieve a fully intelligent closed loop.

[0005] While some industrial systems have deployed operation and maintenance platforms or repair work order systems, most remain at the stage of static work order dispatch and manual review, and have not yet achieved task conflict analysis, priority calculation, and resource collaborative allocation for "corrosion risk + multiple responsibility nodes". Therefore, there is an urgent need to build a multi-dimensional collaborative mechanism based on intelligent analysis of industrial data to realize the allocation of responsibility for corrosion early warning, conflict identification, and optimized task scheduling, providing industrial organizations with a dynamic, controllable, and explainable management solution.

[0006] To address this, a comprehensive safety protection and fault early warning system for cooling towers is proposed. Summary of the Invention

[0007] The purpose of this invention is to provide a comprehensive safety protection and fault early warning system for cooling towers, enabling dynamic hierarchical management of corrosion information, identification of responsibility node conflicts, and adaptive scheduling of response tasks, thereby promoting the transformation of industrial data management from "perception silos" to "closed-loop control." The system includes: a "sound-chemical" data acquisition module, which collects "sound-chemical" datasets; a corrosion risk identification module, which establishes a "sound-chemical" coupling model to assess corrosion risk in target areas and output spatial positioning information; a responsibility node matching module, which automatically matches responsibility nodes and constructs an initial "corrosion point-responsibility node" association; a cross-responsibility identification and optimal response path generation module, which uses a responsibility coverage conflict measurement model to calculate the responsibility coverage conflict degree of each responsibility node and determine whether cross-nodes exist; if cross-nodes exist, a comprehensive scheduling scoring model is constructed, generating an optimal response path based on task response latency, load balancing index, and node priority weight; and a task scheduling module, which, according to the optimal response path, dispatches maintenance tasks to the corresponding responsibility nodes in priority order for hierarchical collaborative response management.

[0008] To achieve the above objectives, the present invention provides the following technical solution:

[0009] A comprehensive safety protection and fault early warning system for cooling towers includes:

[0010] The "sound-chemical" data acquisition module is used to collect "sound-chemical" datasets during the operation of the cooling tower.

[0011] The corrosion risk identification module is used to build a "sound-chemical" coupled model based on the "sound-chemical" dataset, to assess the corrosion risk of the target area, and to output the corresponding spatial positioning information.

[0012] The responsibility node matching module is used to automatically match responsibility nodes and build an initial "corrosion point - responsibility node" relationship based on corrosion risk and spatial location information.

[0013] The cross-responsibility identification and optimal response path generation module is used to calculate the responsibility coverage conflict degree of each responsibility node based on the initial association relationship between "corrosion point - responsibility node" and the responsibility coverage conflict measurement model, and to determine whether there are cross nodes. If there are cross nodes, a comprehensive scheduling scoring model is constructed to generate the optimal response path based on the task response latency, load balancing index and node priority weight of the responsibility node.

[0014] The task scheduling module is used to dispatch maintenance tasks to the corresponding responsible nodes in order of priority based on the optimal response path for hierarchical collaborative response management.

[0015] Preferably, the "acoustic-chemical" dataset includes: vibration noise spectrum data and chemical parameter data;

[0016] The vibration and noise spectrum data includes vibration and acoustic signals at different frequencies during the operation of the cooling tower; the chemical parameter data includes Cl in the water. - Ion concentration, SO4 2- Ion concentration, pH value, conductivity, and dissolved oxygen content.

[0017] Preferably, the acoustic-chemical coupling model includes: an acoustic feature extraction layer, a chemical feature mapping layer, a multimodal feature fusion layer, and a spatial localization and risk assessment output layer;

[0018] The system comprises the following layers: an acoustic feature extraction layer, which performs frequency domain analysis on vibration and noise spectrum data to extract acoustic feature vectors; a chemical feature mapping layer, which calculates corrosion factors based on chemical parameter data, generates chemical feature vectors, and establishes a nonlinear mapping relationship between chemical parameter data and corrosion factors; a multimodal feature fusion layer, which fuses the acoustic and chemical feature vectors to generate a comprehensive corrosion risk feature vector; and a spatial positioning and risk assessment output layer, which divides the cooling tower into N target areas, performs corrosion risk assessment on the comprehensive feature vectors of each target area using a pre-trained multilayer neural network, and outputs the corrosion risk levels and corresponding spatial positioning information for the N target areas.

[0019] Preferably, the process of constructing the initial association between "corrosion point – responsibility node" is as follows:

[0020] A region-responsibility mapping rule base is established, and M responsibility nodes are set for the N target regions of the cooling tower. The responsibility nodes include professional field identifiers, skill levels, equipment configurations, and personnel quantity attributes. An automatic matching algorithm is used to automatically match the corresponding responsibility nodes in the preset region-responsibility mapping rules based on the corrosion risk points of each target region, and the response priority weight is calculated according to the risk level. The corrosion risk points are obtained based on the corrosion risk level and spatial location information. Based on the unique identifier of the corrosion point, its target region, responsibility node identifier, corrosion risk level, spatial location information, response priority, and expected processing time, an association matrix is ​​constructed to obtain the initial association relationship of "corrosion point - responsibility node".

[0021] Preferably, the responsibility coverage conflict measurement model includes: a spatial overlap calculation layer, a temporal conflict analysis layer, a resource competition assessment layer, and a comprehensive conflict measurement layer;

[0022] The spatial overlap calculation layer calculates the percentage of overlapping area of ​​corrosion risk points within the same target area for different responsible nodes based on the initial association between "corrosion point - responsible node," thus obtaining the spatial overlap. The temporal conflict analysis layer calculates the temporal conflict probability based on the task execution time window of each responsible node and the expected processing time in the initial association between "corrosion point - responsible node," thus obtaining the temporal conflict degree. The resource competition assessment layer analyzes the degree of conflict in the demand for the same maintenance resources, equipment configuration, and personnel allocation by different responsible nodes, thus obtaining the resource competition degree. The comprehensive conflict measurement layer synthesizes the responsibility coverage overlap, temporal conflict degree, and resource competition degree to obtain the responsibility coverage conflict degree, and determines whether there are overlapping nodes based on the responsibility coverage conflict threshold.

[0023] Preferably, the comprehensive scheduling and scoring model includes: a task response latency evaluation layer, a load balancing calculation layer, a node priority weight allocation layer, and a path efficiency optimization layer;

[0024] The task response latency assessment layer predicts the task response time of each responsible node and calculates the task response latency based on historical execution data and the current task queue length, combined with the estimated processing time in the initial association between "corrosion point - responsible node". The load balancing calculation layer monitors the current workload rate of each responsible node in real time and calculates the load balancing index based on the distribution of corrosion risk levels in the target area. The node priority weight allocation layer allocates priority weights based on the professional capabilities, equipment configuration, and number of personnel of the responsible nodes, combined with the corrosion risk level of each target area. The path efficiency optimization layer comprehensively considers task response latency, load balancing index, and priority weights to generate the optimal response path and execution order for target areas with overlapping nodes.

[0025] Preferably, the hierarchical collaborative response management process is as follows:

[0026] The urgency of maintenance tasks is classified according to the corrosion risk level of each target area; based on the urgency, corrosion risk level, and priority of the responsible nodes, maintenance tasks are pushed to the corresponding responsible nodes according to the urgency level and the optimal response path is used, and an execution schedule is automatically generated; the system provides functions for receiving and confirming maintenance task receipts, processing progress feedback, and completion status updates, and tracks and records the entire process of maintenance task execution to achieve closed-loop management of maintenance tasks.

[0027] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0028] 1. By constructing a multimodal acoustic-chemical coupling model that integrates acoustic vibration signals and water chemical parameters, the system can achieve early identification and spatial location output of corrosion risks, forming a collaborative response foundation integrating "risk source – spatial location – responsibility node". This acoustic-chemical coupling model significantly improves the confidence level of corrosion point detection, provides stable input for subsequent responsibility node matching and scheduling, and avoids response delays caused by data deviations in downstream links.

[0029] 2. This solution combines corrosion risk with spatial location information to automatically match responsibility nodes, achieving precise location and efficient traceability of corrosion responsibility. This automated initial association between "corrosion point and responsibility node" is fast and accurate, avoiding subjective biases from human judgment. Once corrosion occurs, the system can quickly identify the responsible party, shortening fault response initiation time. The clear chain of responsibility also provides a reliable basis for performance evaluation, maintenance history tracing, and preventative measure development, forming a closed-loop management system to ensure continuous improvement in the quality and efficiency of maintenance work.

[0030] 3. This invention introduces a cross-responsibility node identification module and utilizes a responsibility coverage conflict measurement model to effectively solve problems such as overlapping task responses, ambiguous responsibilities, and resource conflicts among multiple nodes. By constructing a comprehensive scheduling scoring model, the system can combine task response latency, load balancing, and node priority to achieve optimal response path planning with clear responsibilities and reasonable scheduling. This cross-responsibility node identification module is tightly coupled with the initial association module of "corrosion point – responsibility node," making the flow logic of maintenance tasks clearer and significantly improving overall scheduling efficiency and execution accuracy. Attached Figure Description

[0031] Figure 1 This is a schematic diagram of a comprehensive safety protection and fault early warning system for cooling towers provided in an embodiment of the present invention;

[0032] Figure 2 This is a schematic diagram of the "sound-chemistry" coupling model provided in an embodiment of the present invention;

[0033] Figure 3 This is a schematic diagram illustrating the working principle of the responsibility coverage conflict measurement model provided in this embodiment of the invention. Detailed Implementation

[0034] 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.

[0035] This invention proposes a comprehensive safety protection and fault early warning system for cooling towers, which can realize dynamic hierarchical management of corrosion information, identification of responsibility node conflicts, and adaptive scheduling of response tasks, thereby promoting the transformation of industrial data management from "perception silos" to "closed-loop control". To illustrate that the method of this invention can achieve dynamic hierarchical management of corrosion information, identification of responsibility node conflicts, and adaptive scheduling of response tasks, the effectiveness of this invention will be illustrated below with two embodiments.

[0036] Example 1

[0037] In the embodiments of this application, the process of realizing dynamic hierarchical management of corrosion information, identification of conflicting responsibility nodes, and adaptive scheduling of response tasks using the system proposed in this invention is described in detail, thereby promoting the transformation of industrial data management from "perception silos" to "closed-loop control". Figure 1 The specific structural diagram of the system of this invention includes: a "sound-chemical" data acquisition module, a corrosion risk identification module, a responsibility node matching module, a cross-responsibility identification and optimal response path generation module, and a task scheduling module; wherein, the "sound-chemical" data acquisition module collects the "sound-chemical" dataset during the operation of the cooling tower; the corrosion risk identification module establishes a "sound-chemical" coupling model, assesses the corrosion risk of the target area, and outputs spatial positioning information; the responsibility node matching module automatically matches responsibility nodes and constructs an initial association relationship between "corrosion points and responsibility nodes"; the cross-responsibility identification and optimal response path generation module uses a responsibility coverage conflict measurement model to calculate the responsibility coverage conflict degree of each responsibility node and determines whether there are cross nodes; if there are cross nodes, a comprehensive scheduling scoring model is constructed, and an optimal response path is generated based on task response latency, load balancing index, and node priority weight; the task scheduling module, according to the optimal response path, dispatches maintenance tasks to the corresponding responsibility nodes in priority order for hierarchical collaborative response management. The following is based on... Figure 1 The following explanation is provided regarding the content:

[0038] The "sound-chemical" data acquisition module is used to collect "sound-chemical" datasets during the operation of the cooling tower.

[0039] The "acoustic-chemical" dataset includes: vibration noise spectrum data and chemical parameter data;

[0040] The vibration and noise spectrum data includes vibration and acoustic signals at different frequencies during the operation of the cooling tower; the chemical parameter data includes Cl in the water. - Ion concentration, SO4 2- Ion concentration, pH value, conductivity, and dissolved oxygen content.

[0041] Specifically, vibration acceleration sensors are installed on easily corroded parts such as cooling tower fan brackets, packing support beams, and the main frame of the tower body, with a sampling frequency of 10kHz, covering the 0.1Hz-5kHz frequency band.

[0042] Acoustic sensor arrays were deployed at different heights inside the tower to capture acoustic signals during the operation of the cooling tower, with particular attention paid to abnormal sounds in the mid-frequency range of 100Hz-2kHz.

[0043] The collected vibration signals include, but are not limited to: bearing vibration characteristics (10-1000Hz), structural component resonant frequency (50-500Hz), and filler collision noise (500-2000Hz).

[0044] A multi-parameter water quality monitor was installed in the cooling tower's water collection tank to collect real-time data on Cl- ion concentration and SO42- concentration. 2- Ion concentration, pH value, conductivity, and dissolved oxygen content.

[0045] The "acoustic-chemical" data acquisition module in this embodiment achieves coordinated monitoring of physical signals and chemical indicators by simultaneously acquiring vibration noise spectrum data and chemical parameter data, providing multi-dimensional and high-quality basic data support for the overall system. Specifically, the vibration noise spectrum data can reflect the dynamic health status of the cooling tower structural components in real time; when corrosion leads to a decrease in structural stiffness, a characteristic shift in the vibration spectrum will occur. The chemical parameter data provides direct evidence from the corrosion mechanism level, including Cl... - SO4 2- Ion concentration is directly related to corrosion rate, pH value and conductivity reflect the corrosiveness of the environment, and dissolved oxygen content affects the electrochemical corrosion process. Compared with traditional single monitoring methods, this "sound-chemical" combined acquisition method can detect potential corrosion risks 15-20 days in advance, providing a reliable data foundation for subsequent risk identification modules.

[0046] Preferably, the corrosion risk identification module is used to establish a "sound-chemical" coupling model based on the "sound-chemical" dataset, to assess the corrosion risk of the target area, and to output the corresponding spatial positioning information;

[0047] The "acoustic-chemical" coupling model includes: an acoustic feature extraction layer, a chemical feature mapping layer, a multimodal feature fusion layer, and a spatial localization and risk assessment output layer; (Refer to...) Figure 2 ;

[0048] The system comprises the following layers: an acoustic feature extraction layer, which performs frequency domain analysis on vibration and noise spectrum data to extract acoustic feature vectors; a chemical feature mapping layer, which calculates corrosion factors based on chemical parameter data, generates chemical feature vectors, and establishes a nonlinear mapping relationship between chemical parameter data and corrosion factors; a multimodal feature fusion layer, which fuses the acoustic and chemical feature vectors to generate a comprehensive corrosion risk feature vector; and a spatial positioning and risk assessment output layer, which divides the cooling tower into N target areas, performs corrosion risk assessment on the comprehensive feature vectors of each target area using a pre-trained multilayer neural network, and outputs the corrosion risk levels and corresponding spatial positioning information for the N target areas.

[0049] Specifically, the acoustic feature extraction layer performs a fast Fourier transform on the collected vibration noise spectrum data to extract feature frequencies; calculates the peak amplitude and spectral energy distribution parameters of each frequency band, and constructs acoustic feature vectors.

[0050] The corrosion factors in the chemical feature mapping layer include the corrosion activity index, ion concentration ratio, pH deviation, and conductivity change rate; the corrosion activity index is based on Cl... - ion concentration ratio, SO4 2- The result is obtained by weighted fusion of ion concentration ratio, pH deviation (|pH-7|), and conductivity change rate; the fusion weights are obtained by training based on historical data.

[0051] The multimodal feature fusion layer uses an attention mechanism to weightedly fuse acoustic and chemical features, and the fusion weights are obtained through adaptive learning.

[0052] The spatial positioning and risk assessment output layer divides each cooling tower into 20 target areas (N=20), including: 4 tower top fan areas, 8 packing areas, 4 water spray areas, and 4 water collection pool areas. A pre-trained 5-layer fully connected neural network (number of nodes: 100-64-32-16-5) is used for risk assessment, outputting 5 levels of corrosion risk, including Level 1 very low risk, Level 2 low risk, Level 3 medium risk, Level 4 high risk, and Level 5 very high risk. At the same time, spatial positioning information is output: (tower number, area number, 3D coordinates).

[0053] This embodiment proposes an "acoustic-chemical" coupling model, which achieves deep fusion and intelligent analysis of heterogeneous data through a four-layer architecture. It transforms raw physical signals and chemical parameters into quantifiable corrosion risk assessment results and accurately outputs spatial positioning information, laying a technical foundation for precise management of the overall system. Acoustic features can capture abnormal structural vibrations caused by corrosion (such as resonant frequency shifts and increased abnormal noise), while chemical features quantify the electrochemical driving force of corrosion. Deep fusion of these two types of features through a neural network enables the system to issue early warnings in the early stages of corrosion development, achieving a synergistic effect of 1+1>2. Simultaneously, the spatial positioning function divides the cooling tower into N target areas and assesses each one individually, making corrosion risk no longer a general overall judgment but a quantitative result precise to a specific location. This refined risk assessment and positioning supports the precise matching of the responsibility node matching module, ensuring that each corrosion point can find the most suitable responsible party, significantly shortening the response time. Table 1 shows a comparison of the efficiency of different data detection methods for early corrosion risk identification.

[0054] Table 1 Comparison of Early Corrosion Risk Identification Efficiency

[0055] Monitoring methods Corrosion occurrence time System warning time Early warning lead time False alarm rate Chemical parameters 30 days 25 days 5 days 8％ Acoustic vibration 30 days 22 days 8 days 10％ "Sound-Chemistry" Coupling Model 30 days 16 days 14 days 3％

[0056] Preferably, the responsibility node matching module is used to automatically match responsibility nodes and construct an initial association relationship of "corrosion point - responsibility node" based on corrosion risk and spatial location information;

[0057] The process of constructing the initial association between "corrosion point and responsibility node" is as follows:

[0058] A region-responsibility mapping rule base is established, and M responsibility nodes are set for the N target regions of the cooling tower. The responsibility nodes include professional field identifiers, skill levels, equipment configurations, and personnel quantity attributes. An automatic matching algorithm is used to automatically match the corresponding responsibility nodes in the preset region-responsibility mapping rules based on the corrosion risk points of each target region, and the response priority weight is calculated according to the risk level. The corrosion risk points are obtained based on the corrosion risk level and spatial location information. Based on the unique identifier of the corrosion point, its target region, responsibility node identifier, corrosion risk level, spatial location information, response priority, and expected processing time, an association matrix is ​​constructed to obtain the initial association relationship of "corrosion point - responsibility node".

[0059] Specifically, for the 20 target areas, 6 responsibility nodes are set, including the Mechanical Maintenance Group of the Equipment Department, the Electrical Maintenance Group of the Equipment Department, the Operation Management Group of the Production Department, the Water Treatment Workshop, Outsourced Maintenance Unit A, and Outsourced Maintenance Unit B;

[0060] The priority weight P = 0.5 × risk level / 5 + 0.3 × regional importance + 0.2 × historical failure frequency; where regional importance is: fan area (1.0) > packing area (0.8) > water spray area (0.6) > water collection pool area (0.4).

[0061] Table 2 provides a comparison of the effects of different management models on corrosion liability determination and traceability efficiency.

[0062] Table 2 Comparison of Corrosion Liability Determination and Traceability Efficiency

[0063]

[0064] The responsibility node matching module in this embodiment establishes an initial association between "corrosion point" and "responsibility node," achieving automated connection from risk identification to responsibility implementation. This completely solves the problem of unclear responsibilities in traditional management and provides organizational assurance for the efficient operation of the overall system. This module transforms the technical problems of corrosion management into an executable management solution. Through a pre-set region-responsibility mapping rule base, each corrosion point can be automatically matched to the most suitable responsibility node based on its characteristics (location, risk level, professional requirements). The matching process considers multiple dimensions such as professional field, skill level, equipment configuration, and number of personnel, ensuring the scientific nature of the matching. Simultaneously, the calculation of response priority weights allows resources to be prioritized for high-risk areas, avoiding inefficient "all-encompassing" management. This automated matching mechanism greatly improves the efficiency of protection management and provides a clear initial association for the subsequent cross-responsibility identification module, forming the foundation for multi-departmental collaboration.

[0065] Preferably, the cross-responsibility identification and optimal response path generation module is used to calculate the responsibility coverage conflict degree of each responsibility node based on the initial association relationship between "corrosion point - responsibility node" using the responsibility coverage conflict measurement model, and to determine whether there are cross nodes.

[0066] The responsibility coverage conflict measurement model includes: a spatial overlap calculation layer, a temporal conflict analysis layer, a resource competition assessment layer, and a comprehensive conflict measurement layer; (Refer to...) Figure 3 ;

[0067] The spatial overlap calculation layer calculates the percentage of overlapping area of ​​corrosion risk points within the same target area for different responsible nodes based on the initial association between "corrosion point - responsible node," thus obtaining the spatial overlap. The temporal conflict analysis layer calculates the temporal conflict probability based on the task execution time window of each responsible node and the expected processing time in the initial association between "corrosion point - responsible node," thus obtaining the temporal conflict degree. The resource competition assessment layer analyzes the degree of conflict in the demand for the same maintenance resources, equipment configuration, and personnel allocation by different responsible nodes, thus obtaining the resource competition degree. The comprehensive conflict measurement layer synthesizes the responsibility coverage overlap, temporal conflict degree, and resource competition degree to obtain the responsibility coverage conflict degree, and determines whether there are overlapping nodes based on the responsibility coverage conflict threshold.

[0068] Specifically, the spatial overlap calculation layer calculates the coverage overlap of different responsible nodes for corrosion points within the same target area; the spatial overlap is obtained based on the ratio of the overlap area to the total coverage area.

[0069] The time conflict analysis layer analyzes the overlap of task time windows based on the Gantt chart principle, and calculates the time conflict degree based on the ratio of overlapping time to total task time.

[0070] The resource competition assessment layer analyzes the competition for the same maintenance resources, equipment configuration, and personnel allocation, and calculates the resource competition degree based on the ratio of overlapping resource demand to total resource capacity.

[0071] The cross-responsibility identification module in this embodiment proactively identifies and quantifies potential conflicts between responsibility nodes through multi-dimensional responsibility coverage conflict measurement. This provides a decision-making basis for the collaborative optimization of the overall system and effectively avoids the management dilemma of duplicate responses or responsibility vacuums among multiple departments. This cross-responsibility identification module comprehensively assesses responsibility conflicts from three dimensions: spatial overlap, temporal overlap, and resource. Spatial overlap identifies overlapping responsibilities among multiple departments in the same area, avoiding redundant construction; temporal conflict analyzes the timing conflicts of task execution, preventing task backlog during critical periods; and resource competition assesses conflicts in the use of resources such as equipment and personnel, ensuring the rational allocation of resources. Through comprehensive conflict measurement, the system can accurately identify cross-nodes requiring coordination, significantly reducing responsibility conflicts. This proactive conflict identification mechanism not only improves the work efficiency of each responsibility node but also provides an optimized entry point for the comprehensive scheduling scoring model, making it a key link in achieving globally optimal scheduling.

[0072] Preferably, if there are overlapping nodes, a comprehensive scheduling and scoring model is constructed to generate the optimal response path based on the task response latency, load balancing index, and node priority weight of the responsible node.

[0073] The comprehensive scheduling and scoring model includes: a task response latency evaluation layer, a load balancing calculation layer, a node priority weight allocation layer, and a path efficiency optimization layer.

[0074] The task response latency assessment layer predicts the task response time of each responsible node and calculates the task response latency based on historical execution data and the current task queue length, combined with the estimated processing time in the initial association between "corrosion point - responsible node". The load balancing calculation layer monitors the current workload rate of each responsible node in real time and calculates the load balancing index based on the distribution of corrosion risk levels in the target area. The node priority weight allocation layer allocates priority weights based on the professional capabilities, equipment configuration, and number of personnel of the responsible nodes, combined with the corrosion risk level of each target area. The path efficiency optimization layer comprehensively considers task response latency, load balancing index, and priority weights to generate the optimal response path and execution order for target areas with overlapping nodes.

[0075] Table 3 presents a comparison of the efficiency of introducing a cross-responsibility identification mechanism.

[0076] Table 3. Efficiency Comparison of Introducing Cross-Responsibility Identification Mechanism

[0077]

[0078] The integrated scheduling and scoring model in this embodiment generates the optimal response path through a multi-objective optimization algorithm, achieving intelligent scheduling under complex constraints and elevating the system's collaborative management capabilities to a new level, ensuring the maximum utilization of maintenance resources. The integrated scheduling and scoring model balances three key elements: response speed, load balancing, and professional matching. Specifically, task response latency assessment ensures rapid response to urgent tasks, shortening the average response time; load balancing calculation avoids resource waste by preventing some nodes from being overloaded while others are idle, improving resource utilization efficiency; and node priority weights ensure that professionals perform their specialized tasks, improving maintenance quality. Through path efficiency optimization, the system can automatically generate fast and efficient execution plans even in complex situations with intersecting nodes. This intelligent scheduling capability, working closely with the aforementioned conflict identification module, maximizes the collaborative efficiency of multiple departments.

[0079] Preferably, the task scheduling module is used to dispatch maintenance tasks to corresponding responsibility nodes in priority order according to the optimal response path for hierarchical collaborative response management. The hierarchical collaborative response management process is as follows:

[0080] The urgency of maintenance tasks is classified according to the corrosion risk level of each target area; based on the urgency, corrosion risk level, and priority of the responsible nodes, maintenance tasks are pushed to the corresponding responsible nodes according to the urgency level and the optimal response path is used, and an execution schedule is automatically generated; the system provides functions for receiving and confirming maintenance task receipts, processing progress feedback, and completion status updates, and tracks and records the entire process of maintenance task execution to achieve closed-loop management of maintenance tasks.

[0081] The task scheduling module in this embodiment achieves full-process digital control of maintenance tasks from generation to completion through hierarchical collaborative response management. This provides execution assurance and a closed-loop feedback mechanism for the overall system, ensuring that every potential corrosion hazard is addressed promptly and effectively. The task scheduling module transforms the analysis results of all the aforementioned modules into executable maintenance actions. By classifying tasks by urgency level, it ensures the rational allocation of resources and timely response; the automatically generated execution schedule eliminates the arbitrariness and inefficiency of manual scheduling; and the tracking and recording of the entire task process provides valuable historical data to support continuous system optimization. The closed-loop management function, through receiving receipts, progress feedback, and completion confirmation, ensures the controllability and traceability of task execution, improving the operational reliability of the cooling tower.

[0082] The comprehensive safety protection and fault early warning system for cooling towers provided by this invention achieves a fundamental transformation in cooling tower corrosion management from passive response to proactive prevention, from decentralized management to system control, and from experience-based decision-making to data-driven approaches through the organic integration of an acoustic-chemical data acquisition module, a corrosion risk identification module, a responsibility node matching module, a cross-responsibility identification and optimal response path generation module, and a task scheduling module. This innovative system deeply integrates acoustic vibration monitoring with chemical parameter analysis, breaking through the limitations of single monitoring methods and increasing the corrosion early warning accuracy to 92.3%, with early warning time advanced by 15-20 days. Through intelligent automatic responsibility node matching and a multi-dimensional conflict identification mechanism, the responsibility confirmation time is shortened from 1 hour to 5 minutes, and the responsibility conflict incidence rate is reduced by 86.4%. Employing a comprehensive scheduling and scoring model based on task response delay, load balancing, and priority weights, globally optimal resource allocation is achieved, increasing resource utilization efficiency by 67.5% and reducing average response time by 62.8%. Through hierarchical collaborative response and full-process closed-loop management, a complete chain of "monitoring-assessment-decision-execution-feedback" is formed, reducing maintenance costs by 32.0%. The coherent data flow and progressive intelligent decision-making architecture formed by these five modules not only solves the pain points of traditional management such as data silos, unclear responsibilities, and resource waste, but also promotes a systemic transformation of industrial equipment maintenance and management models. It provides enterprises with a replicable paradigm for technology-driven management innovation, which has significant technical value and broad application prospects.

[0083] Example 2

[0084] In Example 1, the method proposed in this invention successfully achieved dynamic hierarchical management of corrosion information, identification of conflicting responsibility nodes, and adaptive scheduling of response tasks, thereby promoting the transformation of industrial data management from "perception silos" to "closed-loop control." To further verify the effectiveness of this invention, safety protection management was also implemented for another cooling tower in this embodiment.

[0085] The "sound-chemical" data acquisition module is used to collect "sound-chemical" datasets during the operation of the cooling tower.

[0086] The "acoustic-chemical" dataset includes: vibration noise spectrum data and chemical parameter data;

[0087] The vibration and noise spectrum data includes vibration and acoustic signals at different frequencies during the operation of the cooling tower; the chemical parameter data includes Cl in the water. - Ion concentration, SO4 2- Ion concentration, pH value, conductivity, and dissolved oxygen content.

[0088] This embodiment provides a cross-domain sensing approach, shifting from "fault data fusion" to "risk factor collaborative modeling." Traditional cooling tower fault monitoring mostly focuses on mechanical indicators such as vibration and temperature rise, lacking early detection of potential corrosion factors. Inspired by the coupling phenomenon of acoustic spectrum evolution and chemical environment, this solution proposes a "sound-chemical" dual-domain data collaborative modeling approach. This model integrates microscopic ionic corrosion trends with macroscopic structural acoustic responses, forming a comprehensive risk detection mechanism oriented towards corrosion precursors. This "cross-domain collaborative data sensing" method provides earlier trigger points and clearer corrosion root cause paths for predictive maintenance.

[0089] Preferably, the corrosion risk identification module is used to establish a "sound-chemical" coupling model based on the "sound-chemical" dataset, to assess the corrosion risk of the target area, and to output the corresponding spatial positioning information;

[0090] The "acoustic-chemical" coupling model includes: an acoustic feature extraction layer, a chemical feature mapping layer, a multimodal feature fusion layer, and a spatial positioning and risk assessment output layer;

[0091] The system comprises the following layers: an acoustic feature extraction layer, which performs frequency domain analysis on vibration and noise spectrum data to extract acoustic feature vectors; a chemical feature mapping layer, which calculates corrosion factors based on chemical parameter data, generates chemical feature vectors, and establishes a nonlinear mapping relationship between chemical parameter data and corrosion factors; a multimodal feature fusion layer, which fuses the acoustic and chemical feature vectors to generate a comprehensive corrosion risk feature vector; and a spatial positioning and risk assessment output layer, which divides the cooling tower into N target areas, performs corrosion risk assessment on the comprehensive feature vectors of each target area using a pre-trained multilayer neural network, and outputs the corrosion risk levels and corresponding spatial positioning information for the N target areas.

[0092] Preferably, the responsibility node matching module is used to automatically match responsibility nodes and construct an initial "corrosion point – responsibility node" association based on corrosion risk and spatial location information; the specific process is as follows:

[0093] A region-responsibility mapping rule base is established, and M responsibility nodes are set for the N target regions of the cooling tower. The responsibility nodes include professional field identifiers, skill levels, equipment configurations, and personnel quantity attributes. An automatic matching algorithm is used to automatically match the corresponding responsibility nodes in the preset region-responsibility mapping rules based on the corrosion risk points of each target region, and the response priority weight is calculated according to the risk level. The corrosion risk points are obtained based on the corrosion risk level and spatial location information. Based on the unique identifier of the corrosion point, its target region, responsibility node identifier, corrosion risk level, spatial location information, response priority, and expected processing time, an association matrix is ​​constructed to obtain the initial association relationship of "corrosion point - responsibility node".

[0094] Preferably, the cross-responsibility identification and optimal response path generation module is used to calculate the responsibility coverage conflict degree of each responsibility node based on the initial association relationship between "corrosion point - responsibility node" using the responsibility coverage conflict measurement model, and to determine whether there are cross nodes.

[0095] The responsibility coverage conflict measurement model includes: a spatial overlap calculation layer, a temporal conflict analysis layer, a resource competition assessment layer, and a comprehensive conflict measurement layer;

[0096] The spatial overlap calculation layer calculates the percentage of overlapping area of ​​corrosion risk points within the same target area for different responsible nodes based on the initial association between "corrosion point - responsible node," thus obtaining the spatial overlap. The temporal conflict analysis layer calculates the temporal conflict probability based on the task execution time window of each responsible node and the expected processing time in the initial association between "corrosion point - responsible node," thus obtaining the temporal conflict degree. The resource competition assessment layer analyzes the degree of conflict in the demand for the same maintenance resources, equipment configuration, and personnel allocation by different responsible nodes, thus obtaining the resource competition degree. The comprehensive conflict measurement layer synthesizes the responsibility coverage overlap, temporal conflict degree, and resource competition degree to obtain the responsibility coverage conflict degree, and determines whether there are overlapping nodes based on the responsibility coverage conflict threshold.

[0097] Preferably, if there are overlapping nodes, a comprehensive scheduling and scoring model is constructed to generate the optimal response path based on the task response latency, load balancing index, and node priority weight of the responsible node; the comprehensive scheduling and scoring model includes: a task response latency evaluation layer, a load balancing calculation layer, a node priority weight allocation layer, and a path efficiency optimization layer;

[0098] The task response latency assessment layer predicts the task response time of each responsible node and calculates the task response latency based on historical execution data and the current task queue length, combined with the estimated processing time in the initial association between "corrosion point - responsible node". The load balancing calculation layer monitors the current workload rate of each responsible node in real time and calculates the load balancing index based on the distribution of corrosion risk levels in the target area. The node priority weight allocation layer allocates priority weights based on the professional capabilities, equipment configuration, and number of personnel of the responsible nodes, combined with the corrosion risk level of each target area. The path efficiency optimization layer comprehensively considers task response latency, load balancing index, and priority weights to generate the optimal response path and execution order for target areas with overlapping nodes.

[0099] Inspired by the common pain points of "overlapping responsibilities and buck-passing" in multi-department collaborative management, this embodiment no longer simply assigns detection tasks to multiple candidate responsibility nodes. Instead, it introduces a "responsibility coverage conflict measurement model," treating responsibility response relationships as a measurable and hierarchical network graph structure for the first time. By calculating conflict degree and cross-level classification, it achieves accurate identification and intervention of the degree of responsibility overlap, providing an intelligent and conflict-resolving path optimization strategy for cross-organizational collaborative task assignment. This reflects an organizational game theory modeling mindset that shifts from "responsibility assignment" to "responsibility conflict identification."

[0100] Preferably, the task scheduling module is used to dispatch maintenance tasks to corresponding responsibility nodes in priority order according to the optimal response path for hierarchical collaborative response management. The hierarchical collaborative response management process is as follows:

[0101] The urgency of maintenance tasks is classified according to the corrosion risk level of each target area; based on the urgency, corrosion risk level, and priority of the responsible nodes, maintenance tasks are pushed to the corresponding responsible nodes according to the urgency level and the optimal response path is used, and an execution schedule is automatically generated; the system provides functions for receiving and confirming maintenance task receipts, processing progress feedback, and completion status updates, and tracks and records the entire process of maintenance task execution to achieve closed-loop management of maintenance tasks.

[0102] This embodiment draws on dynamic scheduling theories such as "bottleneck-first scheduling" and "resource preemptive optimization" in the manufacturing field, proposing a comprehensive scoring mechanism based on cross-node risk priority, task load, and response timeliness. This upgrades traditional manual task assignment decisions to "adaptive task collaborative path planning" supported by real-time computing power. This not only improves task execution efficiency but also enables schedulable optimization for multiple objectives (cost, timeliness, and response quality), building a data-responsibility-resource closed-loop system for equipment lifecycle management. It achieves a paradigm shift from "static task triggering" to "dynamic collaborative scheduling" in intelligent scheduling.

[0103] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A comprehensive safety protection and fault early warning system for cooling towers, characterized in that, include: The "sound-chemical" data acquisition module is used to collect "sound-chemical" datasets during the operation of the cooling tower. The "acoustic-chemical" dataset includes: vibration noise spectrum data and chemical parameter data; The corrosion risk identification module is used to build a "sound-chemical" coupled model based on the "sound-chemical" dataset, to assess the corrosion risk of the target area, and to output the corresponding spatial location information. The responsibility node matching module is used to automatically match responsibility nodes and build an initial "corrosion point - responsibility node" association based on corrosion risk and spatial location information; The cross-responsibility identification and optimal response path generation module is used to calculate the responsibility coverage conflict degree of each responsibility node based on the initial association relationship between "corrosion point - responsibility node" and the responsibility coverage conflict measurement model, and to determine whether there are cross nodes. If there are cross nodes, a comprehensive scheduling scoring model is constructed to generate the optimal response path based on the task response latency, load balancing index and node priority weight of the responsibility node. The task scheduling module is used to dispatch maintenance tasks to the corresponding responsible nodes in order of priority based on the optimal response path for hierarchical collaborative response management.

2. The all-round safety protection and fault early warning system for cooling towers according to claim 1, characterized in that, The "acoustic-chemical" dataset includes: vibration noise spectrum data and chemical parameter data; The vibration and noise spectrum data includes vibration and acoustic signals at different frequencies during the operation of the cooling tower; the chemical parameter data includes Cl in the water. - Ion concentration, SO4 2- Ion concentration, pH value, conductivity, and dissolved oxygen content.

3. The all-round safety protection and fault early warning system for cooling towers according to claim 1, characterized in that, The "acoustic-chemical" coupling model includes: an acoustic feature extraction layer, a chemical feature mapping layer, a multimodal feature fusion layer, and a spatial localization and risk assessment output layer; The system comprises the following layers: an acoustic feature extraction layer, which performs frequency domain analysis on vibration and noise spectrum data to extract acoustic feature vectors; a chemical feature mapping layer, which calculates corrosion factors based on chemical parameter data to obtain chemical feature vectors and establishes a nonlinear mapping relationship between chemical parameter data and corrosion factors; a multimodal feature fusion layer, which fuses the acoustic and chemical feature vectors to generate a comprehensive corrosion risk feature vector; and a spatial positioning and risk assessment output layer, which divides the cooling tower into N target areas, performs corrosion risk assessment on the comprehensive feature vectors of each target area using a pre-trained multilayer neural network, and outputs the corrosion risk level and corresponding spatial positioning information for the N target areas.

4. The all-round safety protection and fault early warning system for cooling towers according to claim 1, characterized in that, The process of constructing the initial association between "corrosion point – responsibility node" is as follows: A region-responsibility mapping rule base is established, and M responsibility nodes are set for the N target regions of the cooling tower. The responsibility nodes include professional field identification, skill level, equipment configuration and personnel quantity attributes. Based on the corrosion risk points of each target region, an automatic matching algorithm is used to automatically match the corresponding responsibility nodes in the preset region-responsibility mapping rules, and the response priority weight is calculated according to the risk level. The corrosion risk points are obtained based on the corrosion risk level and spatial location information. Based on the unique identifier of the corrosion point, its target area, the identifier of the responsible node, the corrosion risk level, the spatial location information, the response priority, and the expected processing time, a correlation matrix is ​​constructed to obtain the initial correlation between "corrosion point and responsible node".

5. The all-round safety protection and fault early warning system for cooling towers according to claim 1, characterized in that, The responsibility coverage conflict measurement model includes: a spatial overlap calculation layer, a temporal conflict analysis layer, a resource competition assessment layer, and a comprehensive conflict measurement layer; The spatial overlap calculation layer calculates the percentage of overlapping area of ​​corrosion risk points within the same target area for different responsible nodes based on the initial association between "corrosion point - responsible node," thus obtaining the spatial overlap. The temporal conflict analysis layer calculates the temporal conflict probability based on the task execution time window of each responsible node and the expected processing time in the initial association between "corrosion point - responsible node," thus obtaining the temporal conflict degree. The resource competition assessment layer analyzes the degree of conflict in the demand for the same maintenance resources, equipment configuration, and personnel allocation by different responsible nodes, thus obtaining the resource competition degree. The comprehensive conflict measurement layer synthesizes the spatial overlap, temporal conflict degree, and resource competition degree to obtain the responsibility coverage conflict degree, and determines whether there are overlapping nodes based on the responsibility coverage conflict threshold.

6. The all-round safety protection and fault early warning system for cooling towers according to claim 1, characterized in that, The comprehensive scheduling and scoring model includes: a task response latency evaluation layer, a load balancing calculation layer, a node priority weight allocation layer, and a path efficiency optimization layer. The task response latency assessment layer predicts the task response time of each responsible node and calculates the task response latency based on historical execution data and the current task queue length, combined with the estimated processing time in the initial association between "corrosion point - responsible node". The load balancing calculation layer monitors the current workload rate of each responsible node in real time and calculates the load balancing index based on the distribution of corrosion risk levels in the target area. The node priority weight allocation layer allocates priority weights based on the professional capabilities, equipment configuration, and number of personnel of the responsible nodes, combined with the corrosion risk level of each target area. The path efficiency optimization layer comprehensively considers task response latency, load balancing index, and priority weights to generate the optimal response path and execution order for target areas with overlapping nodes.

7. The all-round safety protection and fault early warning system for cooling towers according to claim 1, characterized in that, The process of hierarchical collaborative response management is as follows: The urgency of maintenance tasks is classified according to the corrosion risk level of each target area; based on the urgency, corrosion risk level, and priority of the responsible nodes, maintenance tasks are pushed to the corresponding responsible nodes according to the urgency level and the optimal response path is used, and an execution schedule is automatically generated; the system provides functions for receiving and confirming maintenance task receipts, processing progress feedback, and completion status updates, and tracks and records the entire process of maintenance task execution to achieve closed-loop management of maintenance tasks.

Images