Intelligent swimming pool management platform

Abstract

The invention provides a swimming pool intelligent management platform, which is characterized by comprising a partition water quality detection module for dividing a swimming pool into a plurality of detection areas, arranging a water quality sensor in each detection area, collecting water quality parameters in real time and generating water quality parameter data; the visual detection module is used for arranging an image acquisition device in the swimming pool, identifying garbage on the water surface and the bottom of the swimming pool through a visual analysis strategy based on a detection image acquired by the image acquisition device, and generating garbage detection data; the data processing and analyzing module is used for establishing a water quality state and a water quality change trend model of each detection area based on the water quality parameter data, presetting a future water quality state and analyzing a garbage distribution and generation rule based on the garbage detection data; and the intelligent management module is used for generating a comprehensive swimming pool management scheme through an intelligent control strategy according to the future water quality state and the garbage distribution condition and by acquiring the swimming pool utilization rate and the environmental parameters.

Description

A smart pool management platform Technical Field

[0001] This invention relates to swimming pool management, and more particularly to an intelligent swimming pool management platform. Background Technology

[0002] With the increasing awareness of fitness among the public, swimming pools, as important venues for sports, fitness, and leisure, have their management quality directly affecting the health, safety, and experience of users. Currently, the swimming pool management field generally suffers from problems such as traditional management models, low levels of automation, and insufficient precision, making it difficult to meet the needs of large-scale, high-standard swimming pool operations.

[0003] In terms of water quality management, existing technologies mostly employ single-point water quality testing or manual periodic sampling. Single-point testing cannot comprehensively reflect the water quality differences in different areas of the swimming pool, while manual testing is subject to lag and cannot capture dynamic changes in water quality in real time. This leads to operations such as water changes and chemical dosing relying heavily on experience-based judgment. Such blind operations not only easily cause the risk of water quality exceeding standards and threatening the health of users, but also result in excessive water changes and excessive chemical dosing, causing serious waste of water resources and chemicals and increasing operating costs.

[0004] In terms of sanitation and cleaning, the removal of trash from the pool surface and bottom mainly relies on manual inspection and retrieval. Manual trash identification is inefficient, has a high rate of missed detection, and cannot accurately track trash distribution and generation patterns. Furthermore, cleaning path planning lacks scientific basis, often resulting in random operations, leading to cleaning blind spots, increased energy consumption due to repetitive work, and even secondary pollution during the cleaning process. In addition, existing cleaning methods cannot flexibly match cleaning equipment according to the type and quantity of trash, further reducing cleaning efficiency. Summary of the Invention

[0005] In view of the shortcomings of the existing technology, the purpose of this invention is to provide a smart pool management platform to overcome the above-mentioned defects in the existing technology.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a swimming pool intelligent management platform, comprising a zoned water quality detection module, which divides the swimming pool into multiple detection zones, each equipped with a water quality sensor to collect water quality parameters in real time and generate water quality parameter data; a visual detection module, which deploys image acquisition devices within the swimming pool, and based on the detection images acquired by the image acquisition devices, identifies trash on the pool surface and bottom through a visual analysis strategy, and generates trash detection data; a data processing and analysis module, which establishes a water quality status and water quality change trend model for each detection zone based on the water quality parameter data, presets future water quality status, and analyzes the distribution and generation patterns of trash based on the trash detection data; and an intelligent management module, which, based on the future water quality status, trash distribution, and acquisition of pool usage rate and environmental parameters, generates a comprehensive swimming pool management scheme including water exchange areas, water exchange volume, water exchange time, cleaning tasks, and equipment operation modes through an intelligent control strategy. Preferably, the intelligent control strategy includes a differentiated water exchange step, which includes identifying detection areas where the predicted water quality will exceed a preset safety threshold as predicted meter reading areas, selecting either a local water exchange or a circulating replacement method based on the predicted meter reading level, the volume of water in the area, and the circulating filtration capacity, and calculating the required minimum water exchange volume. The local water exchange method involves extracting only the water in the area and replenishing it with an equal amount of new water, while the circulating replacement method involves introducing the water in the area into a circulating filtration system for purification and replenishing it with some new water.

[0007] Preferably, the intelligent control strategy further includes an intelligent dosing step, which includes calculating the required dosage and timing of dosing based on the predicted water quality deviation value of the meter reading area, the water volume of the area, the current concentration of the agent and the target concentration, and generating a dosing control command; the dosing timing is set in advance of the predicted water quality exceedance to ensure that the agent takes effect before the exceedance occurs.

[0008] Preferably, the intelligent control strategy includes a multi-objective optimization step, which includes setting optimization objectives, such as maximizing water saving rate and minimizing water quality identification time. The optimization variables include water exchange area and volume, water exchange method, type and dosage of chemicals, timing of chemical dosing, cleaning task allocation, and equipment operation mode. Constraints are constructed by combining water quality prediction results, waste detection data, pool usage rate, and environmental parameters to generate an optimal solution set. A feasible solution that meets the current management preferences is selected from the optimal solution set, and a comprehensive pool management plan is generated.

[0009] Preferably, the visual analysis strategy includes analyzing the detected image, identifying the garbage targets in the detected image, outputting the category label of each garbage target, converting the image coordinates into three-dimensional spatial coordinates of the pool, calculating the estimated quantity of garbage for each category, and generating garbage information.

[0010] Preferably, the intelligent control strategy includes a path optimization step, which is used to analyze waste information, determine its physical characteristics and corresponding cleaning methods based on waste type, establish a matching relationship between waste type and cleaning equipment, and calculate the cleaning restrictions for each detection area based on waste quantity density, type hazard degree, and location accessibility. Based on the cleaning priority score, combined with the pool map and obstacle information, a multi-objective optimization algorithm is used to generate a cleaning path scheme for each cleaning equipment under the constraints of maximizing coverage, minimizing cleaning time, minimizing equipment energy consumption, and avoiding secondary pollution.

[0011] Preferably, the visual detection module is equipped with a human posture estimation and behavior recognition model to identify the swimmer's dangerous state in real time, generate situation information including the swimmer's location, danger type and danger level, and trigger the corresponding level of rescue dispatch instructions according to preset classification rules.

[0012] Preferably, the pool management module identifies and counts the number of people entering and leaving the pool in real time, generates real-time total number of people in the pool, counts the number of people in each detection area in real time and calculates the personnel density, generates a personnel density map, and sets a safety threshold. It compares the personnel density of each zone with the safety threshold in real time, and generates a flow restriction control command when the safety threshold is exceeded.

[0013] The beneficial effects of this invention are as follows: It achieves precise water quality control, conserves resources, and ensures water safety. This platform uses a zoned water quality detection module to perform real-time, multi-area coverage monitoring of the swimming pool. Combined with a water quality status and trend model constructed by the data processing and analysis module, it can accurately predict future water quality conditions. In conjunction with differentiated water exchange and intelligent chemical dosing steps in the intelligent control strategy, it can selectively implement localized water exchange or circulation replacement in areas predicted to exceed standards. It accurately calculates the minimum water exchange volume, optimal chemical dosage, and timing of chemical dosing, avoiding the blindness of traditional experience-based operations. This not only ensures that chemicals take effect before water quality exceeds standards, guaranteeing stable water quality across the entire area, but also significantly reduces water and chemical consumption, lowering operating costs. It improves cleaning efficiency and quality while reducing operational energy consumption. This platform uses a visual detection module to accurately identify the type, quantity, and three-dimensional spatial location of waste in the swimming pool. Combined with a path optimization step, it can match corresponding cleaning equipment based on waste characteristics and determine cleaning priorities based on waste density and hazard level, planning the optimal cleaning path through a multi-objective optimization algorithm. This design enables precise and orderly cleaning operations, effectively eliminating cleaning blind spots, while maximizing coverage, minimizing cleaning time, reducing equipment energy consumption, and avoiding secondary pollution. Compared with traditional manual cleaning and random operation modes, it significantly improves cleaning efficiency and quality, and reduces labor and energy costs. Attached Figure Description

[0014] Figure 1 is an overall module diagram of the present invention; Figure 2 is a water quality testing flowchart of the present invention; Figure 3 is a visual testing flowchart of the present invention. Detailed Implementation

[0015] 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.

[0016] It should be noted that when a component is described as "fixed to" another component, it can be directly on the other component or may have a component in between. When a component is considered "connected to" another component, it can be directly connected to the other component or may have a component in between. When a component is considered "set on" another component, it can be directly set on the other component or may have a component in between. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.

[0017] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0018] The embodiments of the present invention will be further described in detail below with reference to the accompanying drawings: As shown in Figures 1-3, the present invention provides a smart swimming pool management platform, including a zoned water quality detection module. The pool is divided into multiple detection zones, and water quality sensors are installed in each zone to collect water quality parameters in real time and generate water quality parameter data. Through scientific zone division and precise sensor deployment, comprehensive capture and reliable collection of pool water quality parameters are achieved. The division of detection zones is systematically planned based on the actual structural characteristics of the pool and the differences in usage scenarios. It fully considers the characteristics of water flow circulation paths in the pool's shallow and deep zones, as well as the different densities of personnel activity, ensuring that each detection zone has independent water quality monitoring significance. This avoids situations where excessively large zones prevent the timely detection of local water quality anomalies, or where overly small zones lead to redundant waste of monitoring resources, ultimately forming a zoned system covering the entire pool area without monitoring blind spots. The platform focuses on collecting core parameters closely related to pool water quality safety, including key indicators reflecting water cleanliness and safety such as residual chlorine concentration, pH value, turbidity, and dissolved oxygen content.

[0019] The visual inspection module is equipped with image acquisition devices deployed within the swimming pool. Based on the images captured by these devices, a visual analysis strategy is used to identify debris on the pool surface and bottom, generating debris detection data. The image acquisition devices are high-definition waterproof imaging equipment, requiring underwater pressure resistance, corrosion resistance, and resistance to light refraction interference. They must withstand long-term erosion from disinfectants, water conditioners, and other chemicals in the pool water, while adapting to different lighting environments underwater and on the surface. The deployment plan comprehensively considers the pool's dimensions, depth distribution, and structural layout. Equipment is strategically placed above the pool walls, in key areas of the pool bottom, and at pool corners—locations prone to blind spots—ensuring seamless coverage and complete monitoring of both the surface and bottom. To address insufficient or uneven underwater lighting, low-power supplementary lighting devices are deployed around the imaging equipment. The supplementary lighting intensity can adaptively adjust according to changes in ambient light, avoiding glare from strong light or image blurring from weak light, ensuring the clarity and detail of the captured images. The sampling frequency of the equipment can be dynamically adjusted according to the usage status of the pool. During peak hours when there are many people, the sampling frequency can be increased to accurately capture the garbage generated in an instant; during off-peak hours when there are few people, the frequency can be reduced to reduce invalid data collection and system energy consumption.

[0020] The visual analysis strategy includes analyzing the detected images, identifying litter targets in the images, outputting a category label for each litter target, converting the image coordinates into three-dimensional spatial coordinates of the pool, calculating the estimated quantity of litter for each category, and generating litter information.

[0021] First, the acquired raw images undergo preprocessing. To address issues such as reflections and water wave interference in surface images, and turbidity and color differences in underwater images, image enhancement algorithms are used to improve the contrast between the litter target and the background. Noise reduction algorithms are employed to filter redundant information caused by water ripples and lighting changes, and to repair target contour distortions caused by refraction, ensuring that the features of the litter target are clearly presented. After preprocessing, a target detection algorithm is used to screen for potential litter targets in the images. This algorithm can accurately distinguish litter from non-litter interference items such as swimmers' limbs, pool bottom tile textures, and floating debris based on contour features, texture features, and grayscale differences, eliminating the risk of misidentification.

[0022] For the selected potential waste targets, key physical features are further extracted using feature extraction algorithms, including shape outline, size range, color characteristics, and material-related image features. Based on these features, a classification model is established to categorize the waste targets, clearly distinguishing different types of waste such as leaves, plastic scraps, hair, and paper scraps. Subsequently, by combining the installation location and shooting angle of each image acquisition device with a pre-set 3D spatial model of the swimming pool, a correspondence between image pixel coordinates and the actual spatial location of the pool is established. This converts the 2D coordinates of the waste targets in the image into 3D spatial location information within the pool, accurately locating the specific area where the waste is located.

[0023] In the quantity estimation stage, based on the image size and proportion of the waste targets, as well as the imaging ratio of their respective areas, and combined with the common volume characteristics of similar waste, the quantity estimate of each type of waste is calculated logically. Simultaneously, the distribution density of waste in each detection area is statistically analyzed. The final generated waste detection data includes waste category information, three-dimensional spatial location, quantity estimate, distribution density, and collection timestamp. This data is synchronously transmitted to the data processing and analysis module, providing comprehensive and accurate basic data support for subsequent waste distribution pattern analysis and cleaning task planning.

[0024] The data processing and analysis module establishes models of water quality status and trends in each detection area based on water quality parameter data, and pre-determines future water quality conditions. It also analyzes waste distribution and generation patterns based on waste detection data. After receiving structured water quality parameter data from the zoned water quality detection module, it first performs systematic preprocessing to ensure data quality. To address abnormal data caused by electromagnetic interference or momentary equipment malfunctions during sensor acquisition, a time-series continuity verification method is used. This method compares the fluctuation range of the same parameter in a detection area at adjacent acquisition times, and, combined with the physical characteristics of the parameter, eliminates outliers exceeding reasonable fluctuation ranges. Simultaneously, a combination of mean imputation and trend continuation is used to fill in missing values ​​during data transmission, ensuring the integrity of the data sequence.

[0025] Based on data preprocessing, key features are extracted. These extracted features include the instantaneous values ​​of core water quality parameters in each detection area, the rate of change per unit time, fluctuation amplitude, peak and trough values, and correlations between parameters. For example, the synergistic relationship between residual chlorine concentration and pH value changes is analyzed, and features such as the cumulative rise time of turbidity are extracted. These features comprehensively reflect the dynamic changes of water quality parameters, providing information-rich feature inputs for subsequent model construction. The water quality change trend model is built based on time-series data mining and correlation analysis of influencing factors. Its core logic is to reveal the inherent laws governing the changes of water quality parameters over time and to combine external influencing factors to achieve trend extrapolation. The model first performs time-series analysis on historical water quality parameter data. By analyzing the water quality change trajectories under different time periods, seasons, and pool usage rates, it mines the periodic and trend characteristics of parameter changes, such as the differences in water quality changes between weekdays and weekends, and the rate of water quality deterioration during high-temperature periods in summer.

[0026] Subsequently, the model incorporates external influencing factors such as pool usage and environmental parameters into its analytical framework, establishing the correlation between changes in water quality parameters and these factors. Environmental parameters include temperature, humidity, light intensity, and rainfall, which indirectly affect water quality by influencing processes such as water evaporation, disinfectant decomposition, and microbial reproduction. By analyzing the patterns of water quality changes under different combinations of influencing factors, the model constructs a multivariate trend prediction model.

[0027] The prediction of future water quality is based on a trend model, combining real-time water quality parameter data, predicted pool usage data for a future period, and environmental prediction data. A trend extrapolation algorithm is used to achieve time-segmented predictions. Short-term predictions focus on water quality changes in the next few hours, accurately predicting the trajectory of water quality parameters and their potential state in each monitoring area. Long-term predictions target water quality trends for the next 1 to 3 days, identifying potential areas and time windows where water quality may exceed standards. The prediction results not only include the future water quality status level of each monitoring area but also provide the specific range of changes and key nodes for each parameter, providing a precise basis for the intelligent management module to formulate prevention and control measures.

[0028] The analysis of waste distribution patterns is based on structured waste information generated by the visual detection module, and unfolds from three dimensions: spatial distribution, temporal distribution, and category distribution. Spatial distribution analysis identifies high-distribution and low-distribution areas of waste by statistically analyzing the quantity density and accumulation state of different types of waste in each detection area, and analyzes the causes of distribution differences, such as the tendency of floating waste to accumulate near the pool inlet and the lower distribution of waste at the bottom of the pool in deep water areas. At the same time, it explores the correlation between waste distribution and pool structure, water flow path, and areas of human activity.

[0029] Temporal distribution analysis analyzes the changes in total waste generation and category proportions across different time periods, dates, and seasons to uncover temporal patterns in waste generation. For example, waste generation increases significantly during peak hours when people are densely populated, and the proportion of leaf litter rises during the leaf-fall season. Category distribution analysis focuses on the quantity proportions and regional differences in different types of waste to identify the main types of waste and pollution sources within the pool, providing a basis for matching cleaning equipment and optimizing cleaning strategies. The discovery of waste generation patterns is centered on correlation analysis, combining waste monitoring data with external influencing factors to identify key factors and underlying logic affecting waste generation. First, the correlation between waste generation and pool usage rate is analyzed to quantify the waste generation rate under different personnel densities, clarifying the degree to which the intensity of personnel activity affects waste generation.

[0030] Secondly, the impact of natural factors on waste generation is analyzed in conjunction with environmental parameters. For example, strong winds can easily cause external debris such as leaves and dust to enter the pool, while heavy rainfall may bring surface debris into the pool. Simultaneously, the correlation between waste generation and pool operation and management measures is analyzed, such as whether anti-slip mats are placed at the pool entrance and whether pre-entry cleaning reminders are provided. These measures can affect the amount of waste brought in by human intervention.

[0031] Through multi-dimensional correlation analysis, the core patterns of pool waste generation were finally summarized, including the main sources of waste, peak periods, peak areas, and the primary and secondary relationships of influencing factors. These patterns can provide data support for the intelligent management module to optimize cleaning task allocation and adjust cleaning frequency, enabling forward-looking planning and precise deployment of cleaning operations.

[0032] The intelligent management module, based on future water quality conditions, waste distribution, and data on pool usage and environmental parameters, generates a comprehensive pool management plan through intelligent control strategies. This plan includes information on water exchange areas, water exchange volume, water exchange time, cleaning tasks, and equipment operation modes. First, it gathers valid data from multiple sources, including future water quality conditions (including the risk of exceeding standards in each area and parameter change trends) and waste distribution and generation patterns (type, quantity density, and high-incidence areas) output by the data processing and analysis module; real-time usage rates and zoning personnel density statistics from the pool management module; and environmental parameters such as temperature, light, and wind. Outliers are removed through data verification to ensure a reliable basis for decision-making.

[0033] Based on this data, the intelligent control strategy performs multi-step collaborative calculations: differentiated water exchange identifies and predicts areas exceeding standards, and combines the degree of exceeding standards, water volume, and filtration capacity to select local water exchange or circulation replacement and calculate the minimum water exchange volume; intelligent chemical dosing is linked to water exchange operations, calculating the dosage based on water quality deviation and water volume, and pre-setting the dosing timing to ensure the chemicals take effect; multi-objective optimization focuses on maximizing water conservation and minimizing the time to achieve water quality standards, integrating parameters such as water exchange, chemical dosing, and cleaning to construct constraints and screen feasible solutions that fit management preferences; path optimization matches cleaning equipment based on waste characteristics, determines cleaning priorities based on density and hazard level, and plans efficient cleaning paths.

[0034] The final comprehensive solution clearly marks the water replacement area, water volume, and water replacement time, accurately allocates cleaning tasks, and sets the operating modes of circulation pumps and cleaning equipment. All operations are avoided during peak hours, ensuring water quality safety and cleaning effectiveness while achieving water and chemical conservation and reducing energy consumption, adapting to the real-time operating status of the swimming pool. The intelligent control strategy includes differentiated water replacement steps. These steps identify areas where water quality is predicted to exceed a preset safety threshold as predicted meter reading areas. Based on the predicted meter reading level, the area's water volume, and the circulation filtration capacity, the solution selects between partial water replacement or circulation replacement, calculating the minimum required water replacement volume. Partial water replacement involves extracting only the water in that area and replenishing it with an equal amount of new water. Circulation replacement involves introducing the water in that area into the circulation filtration system for purification and replenishing it with some new water. The accurate identification of areas predicting exceeding standards is based on a future water quality state model output by the data processing and analysis module. It compares the predicted water quality parameters of each detection area with the preset safety threshold, selecting areas where the parameters are highly likely to exceed the threshold as areas requiring treatment. The preset safety thresholds are based on swimming pool water quality safety standards and are formulated in combination with the swimming pool usage scenarios and operational needs to ensure that they meet actual usage requirements.

[0035] Subsequently, key decision-making factors were analyzed. The degree of exceedance was determined by the deviation of the quantitative prediction parameters from the safety threshold. Small deviations indicate slight exceedances, while large deviations indicate severe exceedances. The volume of the regional water body was pre-calibrated based on the size and water depth data during the zoning planning, which is the basis for calculating the water exchange volume. The circulation filtration capacity was determined by real-time monitoring of indicators such as the purification volume per unit time and filtration efficiency of the circulation system, reflecting the current water purification potential of the system.

[0036] The selection of water replacement methods follows the principle of adaptability. If the water quality is slightly substandard and the circulation filtration capacity is sufficient, the circulation replacement method is preferred. The water in this area is introduced into the circulation filtration system for purification treatment. After removing pollutants, it is returned to the original area, and only a small amount of new water generated by evaporation and leakage is added to reduce the consumption of new water. If the water quality is severely substandard or the circulation filtration capacity is insufficient, the local water replacement method is adopted. The substandard water in this area is directly extracted, and an equal amount of new water that meets the standards is added at the same time to quickly improve the water quality.

[0037] The calculation of the minimum water exchange volume revolves around the core objective of achieving water quality standards, and is derived by combining the water body's self-purification capacity with the purification efficiency of the water exchange method. In the circulating replacement mode, the minimum water exchange volume equals the amount of water lost in the area, ensuring the stability of the total water volume. In the local water exchange mode, based on reducing the predicted exceeding parameters to within the safe threshold, the water volume that just meets the standard requirements is calculated by combining the water body's dilution effect and the pollutant degradation law, avoiding resource waste caused by excessive water exchange and achieving a balance between water quality safety and water conservation goals.

[0038] The intelligent control strategy also includes an intelligent dosing process. This process involves calculating the required dosage and timing of dosing based on the predicted water quality deviation value, water volume, current reagent concentration, and target concentration for the meter reading area, and generating dosing control commands. The dosing timing is set a time period before the predicted water quality exceedance occurs to ensure the reagent takes effect before the exceedance occurs. The core calculation basis is clearly defined: the predicted water quality deviation value refers to the difference between the predicted exceedance parameter and the target concentration, directly reflecting the reagent replenishment intensity required to achieve water quality standards; the water volume for the area is a fixed data point calibrated during zoning planning, serving as the basis for dosage calculation; the current reagent concentration is collected in real-time by the zone's water quality sensors to ensure the data accurately reflects the actual water condition; and the target concentration is set with reference to swimming pool water quality safety standards, combined with operational scenario requirements, to ensure the water quality meets usage requirements.

[0039] The calculation of the dosage revolves around the concentration balance logic, with the core objective of increasing the concentration of the agent in the predicted area of ​​exceeding the standard from the current value to the target value. It comprehensively considers the dilution effect of the water volume on the agent, while also taking into account the natural loss and reaction efficiency of the agent in the water body, and derives the dosage of the agent that just meets the water quality standard requirements. This avoids the failure of control due to insufficient dosage, and also prevents waste and secondary impact on water quality caused by excessive dosage.

[0040] Determining the timing of chemical dosing relies on a water quality trend model. First, the time from the current water quality state to the predicted exceedance state is analyzed. Then, combined with the onset time of the selected chemical, the optimal time for advance dosing is calculated. For example, if a certain type of disinfectant requires two hours to take effect, and the water quality is predicted to exceed the standard in four hours, dosing will begin two hours in advance to ensure the chemical reacts fully before the critical exceedance. If a water exchange operation is being performed simultaneously in the area, the dosing timing will be adjusted to after the water exchange is completed to prevent the chemical from being discharged with the water exceeding the standard, thus ensuring the effectiveness of the dosing.

[0041] Ultimately, the module generates precise dosing control commands based on the calculation results, specifying the start-up time, runtime, and chemical release rate of the dosing equipment, thereby achieving automated and precise control of chemical dosing and providing a key guarantee for stable water quality compliance.

[0042] The intelligent control strategy includes a multi-objective optimization step. This step involves setting optimization objectives, including maximizing water saving rate and minimizing water quality calibration time. Optimization variables include water exchange area and volume, water exchange method, chemical type and dosage, timing of chemical dosing, cleaning task allocation, and equipment operation mode. Constraints are constructed by combining water quality prediction results, waste detection data, pool usage rate, and environmental parameters to generate an optimal solution set. From this set, feasible solutions that meet current management preferences are selected, and a comprehensive pool management plan is generated. The core optimization objectives are clearly defined: maximizing water saving rate focuses on reducing fresh water consumption and chemical waste, lowering operating costs, and meeting environmental protection requirements; minimizing water quality calibration time emphasizes quickly eliminating the risk of water quality exceeding standards, ensuring swimmer health and user experience. These two objectives constitute the core optimization direction, while also considering cleaning efficiency and equipment energy consumption control.

[0043] The optimization variables cover all key aspects of pool management. Water change-related variables accurately predict the scope of areas exceeding standards, the necessary water change volume, and the appropriate water change method. Chemical dosing-related variables clarify the type of chemical, the precise dosage, and the timing of early onset of action. Cleaning-related variables rationally allocate cleaning tasks and work sequence. Equipment operation mode regulates the operating power and frequency of circulation pumps and cleaning equipment. All variables are flexibly adapted to the optimization goals.

[0044] The constraints are designed based on actual operational needs and water quality forecasts to ensure that the water quality in each area meets safety standards after the implementation of the plan. Combined with waste monitoring data, the cleaning tasks are required to cover all waste distribution areas without any blind spots. The pool usage rate is taken into account to avoid conflicts between water changes, cleaning operations and peak personnel hours. Environmental parameters are incorporated to adapt to the impact of temperature, light and other factors on water quality changes and equipment operation, while setting upper limits on resource consumption to prevent excessive water changes and chemical additions.

[0045] By employing a multi-objective optimization algorithm to collaboratively process variables and constraints, potential conflicts between different objectives are balanced, generating an optimal solution set containing multiple suitable options. Ultimately, based on the operator's management preferences—if cost control is prioritized, the focus is on achieving the optimal water-saving rate; if user experience is prioritized, the focus is on quickly meeting water quality standards—feasible solutions that meet the specific needs are selected from the solution set and integrated to form a comprehensive swimming pool management solution covering water changes, chemical dosing, cleaning, and equipment operation.

[0046] The intelligent control strategy includes a path optimization step. This step analyzes waste information, determines the physical characteristics of waste based on its type and the corresponding cleaning method, establishes a matching relationship between waste type and cleaning equipment, and calculates cleaning limitations for each detection area based on waste quantity density, type hazard level, and location accessibility. Based on cleaning priority scores, combined with pool map and obstacle information, a multi-objective optimization algorithm is used to generate cleaning path plans for each cleaning device under the constraints of maximizing coverage, minimizing cleaning time, minimizing equipment energy consumption, and avoiding secondary pollution. The strategy deeply analyzes waste information, identifying its physical characteristics based on type. For example, leaves are lightweight and easily float, plastics are hard and difficult to decompose, and hair is long and easily tangled, thus determining the corresponding cleaning methods. For instance, floating waste is removed by surface retrieval, solid waste at the bottom of the pool is vacuumed using negative pressure, and hair is filtered using specialized equipment. Based on this, a specific matching relationship is established between waste type and cleaning equipment to ensure that each type of waste is matched with the optimal cleaning tool.

[0047] The cleaning priority of each detection area is then calculated by comprehensively considering the quantity and density of garbage (the total amount of garbage per unit space), the degree of harm of the type (e.g., garbage that is more likely to cause equipment blockage or water pollution is more harmful), and the accessibility of the location (e.g., the accessibility of pool corners or the area around equipment). The cleaning priority score is obtained by weighting the multi-dimensional indicators, and the area with the higher score is given priority for cleaning operations.

[0048] When planning the path, the entire pool area map and obstacle information are first integrated to identify unavoidable obstacles such as pool bottom pipelines, ladders, and drains. Then, a multi-objective optimization algorithm is used, with constraints including maximizing coverage to ensure no cleaning dead spots, minimizing cleaning time to improve work efficiency, minimizing equipment energy consumption to reduce operating costs, and avoiding secondary pollution to prevent the spread of waste during cleaning, to dynamically plan the operation paths of each cleaning device. The final generated path plan will determine the start and stop positions, travel routes, and operation sequences of each device based on the equipment type and work area, achieving optimal allocation of cleaning resources and maximizing work efficiency.

[0049] The visual detection module incorporates a human posture estimation and behavior recognition model to identify swimmers' dangerous states in real time. It generates situational information including the swimmer's location, type of danger, and danger level, and triggers corresponding rescue dispatch instructions based on preset grading rules. Utilizing a pool-wide image acquisition device, it captures swimmers' limb movements, body features, and movement trajectories in real time. The model employs deep learning algorithms to accurately distinguish between normal swimming movements and dangerous state characteristics. Common danger types include erratic struggling during drowning, prolonged static floating, rapid movements after choking on water, and staying in deep water beyond one's capabilities.

[0050] Simultaneously, the model categorizes danger levels based on the swimmer's movement duration and the degree of postural abnormality. High-risk corresponds to emergencies such as drowning, medium-risk to states requiring immediate intervention such as choking on water or excessive fatigue, and low-risk to behaviors posing safety hazards such as horseplay. The location in the incident information is derived through mapping and conversion between image coordinates and the three-dimensional spatial coordinates of the swimming pool, ensuring accurate positioning.

[0051] The pre-defined grading rules clearly define the corresponding rescue responses for different levels. In a high-risk situation, the highest-level alarm is immediately triggered, and the swimmer's precise location is simultaneously pushed to the rescuer's terminal and the on-site warning equipment is activated. In a medium-risk situation, a key attention instruction is sent to on-site staff to track the swimmer's status in real time. In a low-risk situation, a safety reminder is broadcast to achieve proactive risk control.

[0052] The pool management module identifies and counts the number of people entering and leaving the pool in real time, generating real-time total pool occupancy data. It also separately counts the real-time number of people in each detection area and calculates the personnel density, generating a personnel density map. A safety threshold is set, and the module compares the personnel density of each zone with the safety threshold in real time. When the safety threshold is exceeded, a flow control command is generated. Image recognition technology or sensing devices capture the dynamic movement of people entering and leaving the pool in real time, and two-way verification of statistical data ensures the accuracy of the real-time total number of people. For each detection area, after counting the real-time number of people, the module calculates the personnel density based on the area area, generating an intuitive personnel density map that clearly shows the congestion level of each area.

[0053] The safety thresholds are set based on the pool's design capacity, regional functional differences, and safety management regulations, taking into full account factors such as personnel activity space and rescue accessibility. The module compares the personnel density of each detection area with the safety threshold in real time. When the density of a certain area exceeds the threshold, a flow restriction control command is immediately generated. This is achieved through entrance gate control, on-site broadcast prompts, and area guidance signage to control the influx of people into that area. At the same time, some people are guided to divert to low-density areas to avoid safety issues such as collisions and rescue obstruction caused by excessive crowding, thus ensuring the operational order of the pool and the safety of users.

[0054] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principle of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A smart swimming pool management platform, characterized in that, It includes a zoned water quality testing module, which divides the swimming pool into multiple testing zones. Water quality sensors are installed in each testing zone to collect water quality parameters in real time and generate water quality parameter data. The visual inspection module is equipped with image acquisition devices in the pool. Based on the detection images acquired by the image acquisition devices, it identifies trash on the surface and bottom of the pool through visual analysis strategies and generates trash detection data. The data processing and analysis module establishes models of water quality status and water quality change trends in each detection area based on water quality parameter data, presets future water quality status, and analyzes the distribution and generation patterns of waste based on the waste detection data. The intelligent management module, based on the future water quality status, waste distribution, pool usage rate, and environmental parameters, generates a comprehensive pool management plan that includes water exchange areas, water exchange volume, water exchange time, cleaning tasks, and equipment operation modes through intelligent control strategies.

2. The intelligent swimming pool management platform according to claim 1, characterized in that, The intelligent control strategy includes a differentiated water exchange step, which includes identifying detection areas where the predicted water quality will exceed a preset safety threshold as predicted meter reading areas. Based on the predicted meter reading level, the water volume in the area, and the circulation filtration capacity, the strategy selects to perform either a local water exchange or a circulation replacement method, and calculates the minimum required water exchange volume. The local water exchange method involves extracting only the water in the area and replenishing it with an equal amount of new water, while the circulation replacement method involves introducing the water in the area into a circulation filtration system for purification and replenishing it with some new water.

3. The intelligent swimming pool management platform according to claim 2, characterized in that, The intelligent control strategy also includes an intelligent dosing step, which includes calculating the required dosage and timing of dosing based on the predicted water quality deviation value of the meter reading area, the water volume of the area, the current concentration of the agent and the target concentration, and generating a dosing control command; the dosing timing is set in advance of the predicted water quality exceedance to ensure that the agent takes effect before the exceedance occurs.

4. The intelligent swimming pool management platform according to claim 1, characterized in that, The intelligent control strategy includes a multi-objective optimization step, which includes setting optimization objectives, including maximizing water saving rate and minimizing water quality identification time. The optimization variables are water exchange area and volume, water exchange method, type and dosage of chemicals, timing of chemical dosing, cleaning task allocation and equipment operation mode. Constraints are constructed by combining water quality prediction results, waste detection data, pool usage rate and environmental parameters to generate an optimal solution set. Feasible solutions that meet the current management preferences are selected from the optimal solution set, and a comprehensive pool management plan is generated.

5. The intelligent swimming pool management platform according to claim 1, characterized in that, The visual analysis strategy includes analyzing the detected image, identifying the garbage targets in the detected image, outputting the category label of each garbage target, converting the image coordinates into three-dimensional spatial coordinates of the pool, calculating the estimated quantity of garbage for each category, and generating garbage information.

6. The intelligent swimming pool management platform according to claim 1, characterized in that, The intelligent control strategy includes a path optimization step, which is used to analyze waste information, determine the physical characteristics and corresponding cleaning methods based on waste type, establish a matching relationship between waste type and cleaning equipment, and calculate the cleaning restrictions for each detection area based on waste quantity density, type hazard degree, and location accessibility. Based on the cleaning priority score, combined with the pool map and obstacle information, a multi-objective optimization algorithm is used to generate a cleaning path scheme for each cleaning equipment under the constraints of maximizing coverage, minimizing cleaning time, minimizing equipment energy consumption, and avoiding secondary pollution.

7. The intelligent swimming pool management platform according to claim 1, characterized in that, The visual detection module is equipped with a human posture estimation and behavior recognition model, which can identify the dangerous state of swimmers in real time, generate situation information including swimmer's location, danger type and danger level, and trigger corresponding level of rescue dispatch instructions according to preset classification rules.

8. The intelligent swimming pool management platform according to claim 1, characterized in that, The pool management module identifies and counts the number of people entering and leaving the pool in real time, generates real-time total number of people in the pool, counts the number of people in each detection area in real time and calculates the personnel density, generates a personnel density heat map, and sets a safety threshold. It compares the personnel density of each zone with the safety threshold in real time, and generates a flow restriction control command when the safety threshold is exceeded.

Images