Water quality guarantee management control system of water dispenser and method thereof

Abstract

The present application relates to the technical field of water quality guarantee management and control of water dispensers, and particularly to a water quality guarantee management and control system for water dispensers and a method thereof, which comprises collecting multi-source service parameters including current water quality parameters, average working pressure of a water pump and cumulative running time of a filter; obtaining short-term water quality prediction values and a sudden pollution index through preset deep learning time series prediction model processing; calculating a device health index; determining a composite water quality deviation; determining a system risk index; comparing with a preset threshold to divide a risk level; determining a health water quality trade-off factor in response to the risk level; determining a control target deviation, and combining the control target deviation, the health water quality trade-off factor and the sudden pollution index to calculate and output a final control output; and the present application solves the inherent conflict between short-term purification precision and long-term operation cost.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of water quality guarantee management and control of water dispensers, in particular to a water quality guarantee management and control system for water dispensers and a method thereof. BACKGROUND

[0002] In the field of smart home and public health, water quality guarantee of water dispensers is a key link to ensure user health. Traditional control strategies mainly rely on passive response to threshold value judgment of current water quality parameters. This method lacks foresight and is difficult to effectively respond to instantaneous mutations or sudden pollution events of water source quality.

[0003] At the same time, traditional systems generally ignore the long-term health status and cumulative loss of core components such as filters and water pumps. The system lacks quantitative evaluation of physical loss of equipment during operation, resulting in inherent conflicts between guaranteeing short-term water purification accuracy and maintaining long-term equipment life and reducing operating costs, which easily causes resource waste, high maintenance costs or premature equipment failure. The existing control logic is usually static and lacks a comprehensive decision-making mechanism that integrates AI prediction, equipment health assessment and real-time water quality risk, which cannot dynamically adjust the operation strategy according to the overall risk level of the system. Therefore, how to build a closed-loop management and control method that integrates prediction, evaluation and decision-making to achieve intelligent dynamic balance among guaranteeing water quality safety, prolonging equipment life and optimizing operating costs is a problem that needs to be solved by those skilled in the art. SUMMARY

[0004] To solve the above technical problems, the present application provides a water quality guarantee management and control system for water dispensers and a method thereof. Specifically, the technical solution of the present application is as follows:

[0005] A water quality guarantee management and control method for water dispensers, comprising:

[0006] S1, collecting multi-source service parameters including current water quality parameters, average working pressure of water pump and cumulative running time of filter;

[0007] S2, based on the multi-source service parameters, obtaining short-term water quality prediction value and sudden pollution index through a preset deep learning time series prediction model;

[0008] S3, combining the cumulative running time of the filter and the average working pressure of the water pump, and based on the preset service life and the preset pressure, calculating the equipment health index;

[0009] S4, combining the current water quality parameters and the short-term water quality prediction value to determine the composite water quality deviation;

[0010] S5, combining the composite water quality deviation, the equipment health index and the sudden pollution index, based on the preset limit value and the weight, determining the system risk index;

[0011] S6, dividing to obtain a risk level based on the system risk index and the equipment health index, and comparing with a preset threshold value;

[0012] S7, determining a health water quality trade-off factor in response to the risk level;

[0013] S8, determining a control target deviation according to the current water quality parameter and the preset target water quality, and combining the control target deviation, the health water quality trade-off factor and the sudden pollution index to calculate and output a final control output.

[0014] Preferably, the equipment health index is calculated, comprising:

[0015] determining a filter health item based on a ratio of a cumulative running time of the filter to a preset filter rated reference service life;

[0016] determining a water pump health item based on a ratio of an average working pressure of the water pump to a preset water pump rated pressure;

[0017] weighting and summing the filter health item and the water pump health item to obtain the equipment health index.

[0018] Preferably, the system risk index is determined, comprising:

[0019] dividing the composite water quality deviation by a preset water quality failure limit value to obtain a water quality risk item;

[0020] calculating a loss item of the equipment health index;

[0021] weighting and summing the water quality risk item, the loss item and the sudden pollution index according to a preset weight to obtain the system risk index.

[0022] Preferably, the risk level is divided, comprising:

[0023] determining the risk level as a safe state in response to the system risk index being less than or equal to a preset normal operation threshold value;

[0024] determining the risk level as a high-risk state in response to the system risk index being greater than the normal operation threshold value and less than or equal to a preset failure boundary threshold value;

[0025] determining the risk level as a failure state in response to the system risk index being greater than the failure boundary threshold value, or the equipment health index being less than a preset health failure limit value.

[0026] Preferably, the health water quality trade-off factor is determined, comprising:

[0027] determining the health water quality trade-off factor as a preset economic mode value in response to the risk level being the safe state;

[0028] In response to the risk level being in the high-risk state, a health water quality trade-off factor is calculated based on the equipment health index and the system risk index;

[0029] In response to the risk level being in the failure state, the health water quality trade-off factor is determined as a preset strong purification mode value.

[0030] Preferably, the health water quality trade-off factor is calculated, including:

[0031] Based on the system risk index, a preset normal operation threshold value, and a preset failure boundary threshold value, a normalized risk factor is calculated;

[0032] The loss term of the equipment health index and the normalized risk factor are calculated to obtain the health water quality trade-off factor.

[0033] Preferably, the final control output is calculated and output, including:

[0034] Based on the control target deviation, a proportional feedback control term is generated;

[0035] Based on the sudden pollution index, an AI feedforward gain term is generated;

[0036] Based on the health water quality trade-off factor, a health balance attenuation term is generated;

[0037] The proportional feedback control term, the AI feedforward gain term, the health balance attenuation term, and the preset basic operation output are combined to calculate the final control output, which is output to the actuator.

[0038] A water quality guarantee management control system for a water dispenser, comprising:

[0039] A parameter acquisition module for acquiring multi-source service parameters including current water quality parameters, average working pressure of a water pump, and cumulative running time of a filter;

[0040] A water quality guarantee management module, comprising:

[0041] A prediction processing unit for acquiring short-term water quality prediction values and a sudden pollution index by processing the multi-source service parameters based on a preset deep learning time series prediction model;

[0042] A health calculation unit for combining the cumulative running time of the filter and the average working pressure of the water pump, and calculating an equipment health index based on a preset service life and a preset pressure;

[0043] A deviation determination unit for determining a composite water quality deviation by combining the current water quality parameters and the short-term water quality prediction values;

[0044] A risk calculation unit for determining a system risk index based on preset limits and weights by combining the composite water quality deviation, the equipment health index, and the sudden pollution index.

[0045] a risk grade division unit configured to divide the risk grade based on the system risk index and the equipment health index and in comparison with a preset threshold value;

[0046] a strategy generation unit configured to determine a health water quality trade-off factor in response to the risk grade;

[0047] a control output unit configured to determine a control target deviation based on the current water quality parameter and the preset target water quality, and calculate and output a final control output in combination with the control target deviation, the health water quality trade-off factor and the sudden pollution index.

[0048] Compared with the prior art, the present application has the following beneficial effects:

[0049] 1. The present application introduces a deep learning model, which can predict short-term water quality and sudden pollution risk; this enables the system to change from passive response to forward-looking intervention, effectively dealing with instantaneous mutations of water sources and improving the timeliness of water quality protection;

[0050] 2. The present application quantifies the health status of core components by evaluating the cumulative running time of the filter and the average working pressure of the water pump; this design converts intangible equipment depreciation into a calculable state variable, solving the problem of ignoring equipment wear and tear in traditional methods;

[0051] 3. The present application integrates water quality risk, equipment health wear and tear and sudden pollution prediction to build a comprehensive system risk index; this multi-dimensional evaluation overcomes the one-sidedness of traditional single threshold judgment, enabling the system to more accurately recognize the real risk level it is in;

[0052] 4. The present application dynamically adjusts the control strategy based on the comprehensive risk grade; the system can prioritize energy saving and equipment life when safe, fully guarantee water quality when failed, and intelligently trade off when at high risk, solving the inherent conflict between short-term purification accuracy and long-term operation cost. BRIEF DESCRIPTION OF DRAWINGS

[0053] The present application will be further explained below in conjunction with the accompanying drawings and examples:

[0054] Figure 1 is a flowchart of the method of the present application;

[0055] Figure 2 is a structural diagram of the system of the present application. DETAILED DESCRIPTION

[0056] To make the purpose, technical scheme and advantages of the present application clearer, further detailed description will be made below in conjunction with specific examples.

[0057] Example 1:

[0058] Referring to Figure 1 A water quality guarantee management control method of a water dispenser, comprising:

[0059] S1, collecting multi-source service parameters including current water quality parameters, water pump average working pressure and filter cumulative running time;

[0060] S2, based on the multi-source service parameters, obtaining short-term water quality prediction values and sudden pollution indexes through a preset deep learning time series prediction model processing;

[0061] S3, combining the filter cumulative running time and the water pump average working pressure, and based on a preset service life and a preset pressure, calculating a device health index;

[0062] S4, combining the current water quality parameters and the short-term water quality prediction values, determining a composite water quality deviation;

[0063] S5, combining the composite water quality deviation, the device health index and the sudden pollution index, based on preset limits and weights, determining a system risk index;

[0064] S6, based on the system risk index and the device health index, and comparing with a preset threshold, dividing to obtain a risk level;

[0065] S7, in response to the risk level, determining a healthy water quality trade-off factor;

[0066] S8, determining a control target deviation according to the current water quality parameters and a preset target water quality, and combining the control target deviation, the healthy water quality trade-off factor and the sudden pollution index, calculating and outputting a final control output.

[0067] The embodiment discloses a water quality guarantee management control method of a water dispenser, which is the core of the technical scheme of the present application, and aims to build a closed-loop management system integrating prediction, evaluation, decision and control; the method comprises the following steps:

[0068] S1, collecting multi-source service parameters including current water quality parameters, water pump average working pressure and filter cumulative running time;

[0069] The multi-source service parameters refer to multi-dimensional data representing the real-time running state and historical cumulative loss of the water dispenser system; in the embodiment, the current water quality parameters are defined as , which are collected in real time by a water quality sensor and can include one or more indicators such as turbidity, total dissolved solids TDS, residual chlorine, etc.; the water pump average working pressure is defined as , which is collected by a pressure sensor and reflects the instantaneous load state of the water pump; the filter cumulative running time is defined as , record the cumulative usage degree of core consumables through the built-in timer; the purpose of this step is to provide necessary and high-quality real-time input data for subsequent prediction and evaluation models;

[0070] S2, based on multi-source service parameters, through a preset deep learning time series prediction model, obtain short-term water quality prediction values and sudden pollution indexes;

[0071] The preset deep learning time series prediction model refers to a model that is trained in advance based on historical water quality data and typical pollution event data, such as a recurrent neural network (RNN). The model receives the multi-source service parameters collected in S1 as input. For example, the model can use historical time series data to predict , and use the mutation characteristics of and to jointly determine ; the short-term water quality prediction value, defined as , is a quantitative prediction of key water quality indicators such as TDS within a short time window in the future; the sudden pollution index, defined as , is a probability value between 0 and 1 output by the model, representing the possibility of the system encountering sudden events such as water source instantaneous pollution; the innovation of this step lies in the introduction of forward-looking AI prediction, which enables the system not only to respond to the current state, but also to intervene in advance of future risks;

[0072] S3, combine the cumulative running time of the filter and the average working pressure of the water pump, and based on the preset service life and the preset pressure, calculate the equipment health index;

[0073] The equipment health index, defined as , is a key quantitative indicator that aims to evaluate the cumulative wear and tear and long-term health of core components such as filters and water pumps; the preset service life refers to the rated reference service life of the filter, defined as , which is derived from the equipment specification; the preset pressure refers to the rated pressure of the water pump, defined as , which is also derived from the equipment specification; this step converts the intangible equipment depreciation into a calculable state variable by comparing the real-time cumulative parameters and the average parameters with the rated benchmarks , , providing a basis for subsequent risk assessment and strategy trade-off;

[0074] S4, determine the composite water quality deviation based on the current water quality parameters and the short-term water quality prediction values;

[0075] The composite water quality deviation, defined as , the purpose is to comprehensively evaluate the current water quality state and the AI predicted future water quality risk; in this embodiment, the deviation is determined by taking the maximum value of the current water quality parameter and the short-term water quality prediction value , that is, taking the worst case; that is ; the purpose of this design is to ensure that the system can respond to both the deterioration that has occurred and the deterioration that is predicted, embodying the defensive and forward-looking design principle;

[0076] S5, combined with the composite water quality deviation, the equipment health index and the sudden pollution index, based on the preset limit value and the weight, determine the system risk index;

[0077] The system risk index, defined as , is one of the core indicators of the present application, and its purpose is to create a single, dimensionless index to quantify the overall degree of the system approaching the high-risk steady state and the failure boundary in real time; the preset limit value refers to the water quality failure limit, defined as , which defines the failure boundary that the water quality parameter can allow, such as the maximum allowed value of TDS; the preset weight refers to the weighting coefficient used to balance different risk sources, water quality, health, and suddenness; this step combines three completely different dimensions, namely, the real-time water quality risk , long-term health loss and probabilistic sudden event , into a unified risk measure;

[0078] S6, based on the system risk index and the equipment health index, and compared with the preset threshold, the risk level is divided;

[0079] The risk level is a qualitative division of the current operating state of the system, and its purpose is to provide clear decision-making basis for the subsequent adaptive strategy selection; the preset threshold includes the normal operating threshold, defined as , the failure boundary threshold, defined as , and the health failure limit, defined as ; in this embodiment, the risk level is divided into safe state, high-risk state and failure state; the key of this step is that it not only depends on the comprehensive system risk index , but also introduces an independent judgment of the equipment health index , which constitutes a double insurance;

[0080] S7, in response to the risk level, determine the health water quality trade-off factor;

[0081] The health water quality trade-off factor, defined as This is a strategy adjustment parameter designed to dynamically address the dilemma between short-term water quality accuracy and long-term equipment health; the system automatically selects different operating modes based on the risk levels defined by S6; for example, in a safe state, It is set to a preset economic mode value, prioritizing energy saving and equipment lifespan; in a failure state... It is set to a preset high-purification mode value to protect water quality at all costs; however, in high-risk states, It will be generated dynamically;

[0082] S8 determines the control target deviation based on the current water quality parameters and the preset target water quality, and calculates and outputs the final control output by combining the control target deviation, the healthy water quality trade-off factor and the sudden pollution index.

[0083] This step is the final execution of the control closed loop; the preset target water quality is defined as... The target water quality standard is the standard that the system should achieve, which can be set by the user or based on industry standards; the deviation from the control target is defined as... This is the current water quality parameter. and The error between them, i.e. The final control output is defined as follows: This refers to the actual control commands, such as power or speed, sent to actuators like water pumps or purification units; the innovation of this step lies in the fact that it is not simply based on deviation. Instead of controlling, the strategy factors determined by S7 are used. The predicted index obtained from S2 It is integrated into the control law to achieve adaptive and forward-looking regulation;

[0084] This embodiment constructs a complete technical closed loop, from multi-source sensing S1 to AI prediction S2, then to comprehensive health and risk assessment S3-S5, followed by hierarchical strategy decision-making S6-S7, and finally achieving fusion control output S8; compared to existing technologies that only base on current water quality... This invention employs a simple threshold control strategy, enabling proactive responses to sudden pollution events. Quantify and balance long-term equipment wear and tear. And based on comprehensive risks Dynamically adjust control strategy This achieves an unprecedented dynamic balance between ensuring water quality safety, extending equipment lifespan, and reducing operating costs, significantly improving the intelligence, reliability, and economy of the water dispenser system.

[0085] Example 2:

[0086] The calculated equipment health index includes:

[0087] determining a filter health item based on a ratio of a filter cumulative running time to a preset filter rated reference life;

[0088] determining a water pump health item based on a ratio of a water pump average working pressure to a preset water pump rated pressure;

[0089] weighting and summing the filter health item and the water pump health item to obtain a device health index.

[0090] The embodiment is a further limitation on the specific implementation of step S3 in embodiment 1, which calculates the device health index. The purpose is to provide a quantifiable device health index based on a physical wear model calculation method;

[0091] In this embodiment, the device health index is calculated, including:

[0092] determining a filter health item based on a ratio of a filter cumulative running time to a preset filter rated reference life; and determining a water pump health item based on a ratio of a water pump average working pressure to a preset water pump rated pressure;

[0093] The filter health item is defined in this embodiment as , and the water pump health item is defined as ;

[0094] weighting and summing the filter health item and the water pump health item to obtain a device health index.

[0095] In order to quantify the implicit cost, i.e. the device health degree, and create a calculable index, the calculation formula of the device health index in this embodiment is as follows:

[0096] ;

[0097] is a system comprehensive health index, which physically represents the comprehensive remaining life or health status of the core components of the system, and is a dimensionless value between 0 and 1, calculated by the formula; and are the health weights of the filter and the water pump, respectively, which are dimensionless values, and their sources are preset according to the component replacement cost and the importance to the system, and satisfy ; is the filter cumulative running time, which is recorded by the timer in step S1; is the filter rated reference life, which is obtained from the device specification, and has the same time unit as ; the ratio Reflects the percentage of the used life of the filter; The average working pressure of the water pump, which is obtained from the pressure sensor in step S1; The rated pressure of the water pump, which is obtained from the equipment specification, and Have the same pressure unit; the ratio Reflects the average load degree of the water pump;

[0098] The technical motivation of this formula is that it normalizes two different dimensions of physical loss, the time loss of the filter and the pressure loss of the water pump, by dividing by the rated value, and converts it into a residual health degree in the form of , by weighting , Fuses into a single, decision-making device health index ; those skilled in the art can understand that And Is a linear simplified model of health degree, and in other embodiments, a nonlinear decay function based on a physical model, such as an exponential function or an S-shaped function, can also be replaced to more accurately represent the actual loss curve of the component, which is within the scope of the present application;

[0099] This embodiment concretizes and implements the device health index in embodiment 1 by providing an explicit mathematical model; compared to the general description calculated in embodiment 1, it provides a quantitative evaluation method based on a physical model, rather than a simple empirical rule or black box model; the formula Converts the abstract device health into a specific, calculable, and real physical loss reflecting engineering parameter, making the system's cognition of its own state more accurate, and providing a solid data foundation for subsequent risk assessment S5 and strategy trade-off S7.

[0100] Embodiment 3:

[0101] Determine the system risk index, including:

[0102] Divide the composite water quality deviation by the preset water quality failure limit to obtain the water quality risk term;

[0103] Calculate the loss term of the device health index;

[0104] Weighted sum the water quality risk term, the loss term, and the sudden pollution index according to the preset weight to obtain the system risk index.

[0105] This embodiment is a further limitation on the specific implementation of embodiment 1, step S5, determining the system risk index; its purpose is to provide a system risk index that integrates the three dimensions of water quality, health, and sudden events quantitative calculation method of the system risk index;

[0106] In this embodiment, the system risk index is calculated, including:

[0107] The composite water quality deviation is divided by the preset water quality failure limit value to obtain a water quality risk term; the water quality risk term is defined in this embodiment as ;

[0108] A loss term of the equipment health index is calculated; the loss term is defined in this embodiment as , which is the complement of the equipment health index, representing the degree of loss;

[0109] The water quality risk term, the loss term, and the sudden pollution index are weighted and summed according to preset weights to obtain the system risk index;

[0110] In order to create a single index that comprehensively reflects water quality stability, equipment health loss, and dynamic interference, the calculation formula of the system risk index in this embodiment is as follows:

[0111] ;

[0112] The system comprehensive risk index, which quantifies the degree of the system approaching the failure boundary, is a dimensionless value and is calculated by the formula; 、 、 are preset adjustable weight coefficients, which are dimensionless values; in order to ensure the implementability of those skilled in the art, the specific sources thereof can be calibrated in the following way: a calibration data set containing historical water quality, equipment loss, and sudden events is collected, and based on the data set, the optimal weight coefficients are fitted by methods such as least squares regression analysis or machine learning optimization algorithm, with the goal of minimizing the historical risk misjudgment rate; in runtime, these calibrated coefficients are fixed in the model; and the sum of the coefficients satisfies ; The composite water quality deviation is determined by step S4, as ; The preset water quality failure limit value has the same water quality unit as , and its source is preset according to safety standards or engineering experience; the water quality risk term is a dimensionless normalized value representing the percentage of the current water quality risk reaching the failure limit value; The equipment health index is calculated by step S3, and the loss term is a dimensionless value representing the accumulated loss degree of the equipment; The sudden pollution index is a dimensionless probability value of 0 to 1, which is obtained by the deep learning model in step S2;

[0113] It normalizes the three different risk sources, namely the instant / predicted water quality risk, the accumulated physical loss risk, and the probabilistic sudden event risk, into one evaluation system through weighted summation;

[0114] The embodiment provides a specific mathematical model to make the system risk index in embodiment 1 specific and implementable; compared with the general description determined in embodiment 1, the gain effect of the embodiment is that it provides a multi-dimensional risk fusion quantitative evaluation method; the formula The innovation of the formula is that the three heterogeneous indexes of water quality, health, and AI prediction are normalized and weighted fused; this enables the system to comprehensively and dynamically evaluate the real risk level, avoiding the one-sidedness brought by a single index such as only looking at water quality , and provides more reliable and robust decision basis for subsequent risk level division S6.

[0115] Embodiment 4:

[0116] The division obtains a risk level, including:

[0117] In response to the system risk index being less than or equal to a preset normal operation threshold, the risk level is determined as a safe state;

[0118] In response to the system risk index being greater than the normal operation threshold and less than or equal to a preset failure boundary threshold, the risk level is determined as a high-risk state;

[0119] In response to the system risk index being greater than the failure boundary threshold, or the device health index being less than a preset health failure limit, the risk level is determined as a failure state.

[0120] The embodiment is a further limitation on the specific implementation of the risk level division in step S6 in embodiment 1; the purpose is to provide a level division rule based on the system risk index and the device health index with clear logic and thresholds;

[0121] In the embodiment, the risk level division rule includes:

[0122] In response to the system risk index being less than or equal to a preset normal operation threshold, the risk level is determined as a safe state;

[0123] The preset normal operation threshold is defined as , which is preset according to simulation or empirical data, such as ;

[0124] The logical condition of this rule is: ; at this time, the system is stable, and the economic mode is activated;

[0125] In response to the system risk index being greater than the normal operation threshold and less than or equal to the preset failure boundary threshold, the risk level is determined to be a high-risk state;

[0126] The preset failure boundary threshold is defined as , which is preset according to simulation or empirical data, such as , and ;

[0127] The logical condition of this rule is: ; at this time, the system enters a high-risk steady state, and the balance mode needs to be activated for adaptive adjustment;

[0128] In response to the system risk index being greater than the failure boundary threshold, or the device health index being less than the preset health failure limit, the risk level is determined to be a failure state;

[0129] The preset health failure limit is defined as , which is preset according to safety standards or engineering experience, such as ;

[0130] The logical condition of this rule is: or ; at this time, the system reaches the failure boundary or is close to physical failure, and the strong purification mode or alarm must be activated;

[0131] This division logic sets a double trigger condition for the failure state; the conventional trigger is based on the comprehensive risk exceeding the failure boundary ; and the non-obvious trigger is to add a forced judgment;

[0132] This embodiment provides a specific, double-verified logical rule to specific the risk level division in embodiment 1; compared with the general description obtained by dividing in embodiment 1, the gain effect of this embodiment is that it provides a more robust and safer risk level division mechanism; in particular, the or logic of the failure state is a key non-obvious improvement; it ensures that even in the case of a acceptable comprehensive risk index The bottom line is reached, such as a filter or water pump is close to damage, the system will also be forced to upgrade to the highest risk level; this avoids the risk of physical failure being masked by the momentary good water quality, greatly improving the system's failure warning ability and operational safety.

[0133] Embodiment 5:

[0134] The health water quality trade-off factor is determined, including:

[0135] In response to the risk level being in a safe state, the health water quality trade-off factor is determined as a preset economic mode value;

[0136] In response to the risk level being in a high-risk state, the health water quality trade-off factor is calculated and generated based on the equipment health index and the system risk index;

[0137] In response to the risk level being in a failure state, the health water quality trade-off factor is determined as a preset strong purification mode value.

[0138] This embodiment is a further limitation of the specific implementation of step S7 in Embodiment 1, which determines the health water quality trade-off factor; the purpose is to match different control strategies according to the risk level divided in Embodiment 4;

[0139] In this embodiment, the health water quality trade-off factor is determined according to the following rules:

[0140] In response to the risk level being in a safe state, the health water quality trade-off factor is determined as a preset economic mode value;

[0141] The preset economic mode value is defined as , which is a preset constant; in this mode, the system prioritizes energy optimization and equipment health; according to the subsequent control law, contains , the larger , the smaller the control output; therefore, is set to a higher value, such as close to 1, to suppress unnecessary purification actions and achieve energy saving and life extension;

[0142] In response to the risk level being in a failure state, the health water quality trade-off factor is determined as a preset strong purification mode value;

[0143] The preset strong purification mode value is defined as , which is a preset constant; in this mode, the system prioritizes water quality stability at any cost; in order to maximize the control output , is set to a minimum value, i.e. ; this makes the health balance decay term becomes 1, the control system is in full force to purify;

[0144] In response to the risk level being in a high-risk state, a health water quality trade-off factor is calculated based on the equipment health index and the system risk index;

[0145] This is the activation of the balance mode; in this state, is no longer a preset constant, but needs to be calculated according to the current and dynamic generation; the purpose is to balance the conflict between water quality stability and equipment health in real time and dynamically when the system is in the sensitive interval of the high-risk steady state;

[0146] The embodiment defines a mode switching strategy based on the risk level, and specifically implements the strategy generation in Embodiment 1; compared with the general description determined in Embodiment 1, the gain effect of the embodiment is that it provides a structured and divide-and-conquer control idea; instead of trying to cover all working conditions with an extremely complex continuous function, the application first divides the system into three clear macro states: safe, high-risk, and failure, and specifies simple and robust preset strategies for the safe, economic priority, and failure, water quality priority, two extreme cases and ; this design greatly simplifies the control logic, and the most complex and fine trade-off calculation is limited to the high-risk intermediate state; the division of this strategy makes the system response clearer and more stable, and easy to debug and verify.

[0147] Embodiment 6:

[0148] The health water quality trade-off factor is calculated, including:

[0149] Based on the system risk index, the preset normal operation threshold, and the preset failure boundary threshold, a normalized risk factor is calculated;

[0150] The loss term of the equipment health index and the normalized risk factor are calculated to obtain the health water quality trade-off factor.

[0151] This embodiment is a further limitation of the specific implementation of calculating the health water quality trade-off factor in the high-risk state in Embodiment 5; the purpose is to provide a nonlinear calculation model that can dynamically balance health loss and water quality risk; in this embodiment, the calculation of the health water quality trade-off factor includes:

[0152] Based on the system risk index, the preset normal operation threshold, and the preset failure boundary threshold, a normalized risk factor is calculated; the normalized risk factor is defined as The calculation formula is as follows: ;

[0153] The technical motivation behind this factor is that it will In high-risk areas The absolute value within is mapped to Relative position within the interval; when near hour, Approaching 0; when near hour, Close to 1;

[0154] The loss component of the equipment health index is calculated together with the normalized risk factor to obtain the healthy water quality trade-off factor.

[0155] The wear and tear component of the equipment health index is... ;

[0156] In order to design a dynamic trade-off mechanism, the healthy water quality trade-off factor in this embodiment The calculation formula is as follows:

[0157] ;

[0158] Right now: ;

[0159] The health-water quality tradeoff factor is a dimensionless value between 0 and 1, calculated using this formula. The equipment health index is obtained from step S3; The lower the health level, the better. The smaller the value, the larger this value, and the more the system tends to protect health. The larger; The system risk index is derived from the calculation in step S5; and These are the normal operation threshold and the failure boundary threshold, respectively, which are derived from the thresholds preset in step S6; The function ensures Always a non-negative value;

[0160] With health loss The worse the health, the more inclined one is to protect equipment. The larger; and normalization risk Inversely proportional, through Item; when risk Approaching the failure boundary hour, tending towards 1, leading to tends to 0, and thus tends to 0; this formula has good robustness; for example, when the system risk is low, the is negative, greater than 1, at which time is still dominated by , as expected by the economic model. At the same time, when becomes negative due to over-service, the term will be greater than 1, causing to further increase, thereby forcibly suppressing the control output to protect the equipment that is close to failure;

[0161] This embodiment implements the balance model in Embodiment 5 by providing a specific, nonlinear mathematical model; compared to the summary description calculated in the previous embodiment, it provides a delicate, nonlinear dynamic trade-off mechanism; the non-obvious improvement of the formula is that it is not a simple linear balance; when the risk is low, the is small, and the health factor is dominant; but when the risk approaches the failure boundary, tends to 1, and the risk factor will overwhelmingly defeat the health factor, forcibly returning to zero to ensure that the system fully guarantees water quality when it is close to failure; this nonlinear characteristic of automatically switching the dominant factor at different risk positions is the key to achieving intelligent trade-off.

[0162] Embodiment 7:

[0163] calculates and outputs the final control output, including:

[0164] generating a proportional feedback control term based on the control target deviation;

[0165] generating an AI feedforward gain term based on the sudden pollution index;

[0166] generating a health balance attenuation term based on the health water quality trade-off factor;

[0167] combining the proportional feedback control term, the AI feedforward gain term, the health balance attenuation term, and the preset basic operation output to calculate the final control output, and outputting the final control output to the actuator.

[0168] The embodiment is a further limitation of the specific implementation of the calculation and output of the final control output in step S8 in embodiment 1; the purpose is to provide a hybrid control law that combines feedback control, AI feedforward, and health balance;

[0169] In this embodiment, the calculation of the final control output includes:

[0170] Based on the control target deviation, a proportional feedback control term is generated;

[0171] The control target deviation ;

[0172] The proportional feedback control term is defined in this embodiment as ; wherein The purpose of the function is to ensure that only when , i.e. the water quality is worse than the target, positive control is applied; if the water quality meets or is better than the target, , then this term is 0, and no additional purification is performed;

[0173] Based on the sudden pollution index, an AI feedforward gain term is generated;

[0174] The AI feedforward gain term is defined in this embodiment as ; when the AI predicts sudden pollution, , this term is greater than 1, amplifying the basic control term and achieving early intervention;

[0175] Based on the health water quality trade-off factor, a health balance attenuation term is generated;

[0176] The health balance attenuation term is defined in this embodiment as ; when the system tends to protect health, , this term is less than 1, suppressing the control output to extend the service life at the cost of part of the water quality accuracy;

[0177] Combining the proportional feedback control term, the AI feedforward gain term, the health balance attenuation term, and the preset basic operation output, the final control output is calculated by fusion;

[0178] In order to fuse the above terms in one control law, the calculation formula of the final control output in this embodiment is as follows:

[0179] ;

[0180] The final control output is the instruction sent to the actuator, such as the water pump, with the unit of power W or frequency Hz, which is calculated by the formula; is the health water quality trade-off factor, dimensionless, source is determined by step S7; is the control target deviation, in water quality units such as ppm, source is ; is the proportional gain, dimension of, control output unit / water quality unit, source is an adjustable parameter determined by system simulation or on-site commissioning; is the sudden pollution index, dimensionless, source is obtained by step S2; is the AI feedforward gain coefficient, dimensionless, source is an adjustable parameter determined by system simulation or on-site commissioning; is the basic operation output, same unit as , physical meaning is the minimum maintenance power consumption or basic output, source is an adjustable parameter determined by system simulation or on-site commissioning;

[0181] It combines the classic proportional feedback as the control base, innovatively uses the health balance and AI feedforward as two multiplicative adjustment factors, and dynamically schedules their gains;

[0182] This embodiment implements the control output calculation in embodiment 1 by providing a specific, multi-factor integrated mathematical model; compared with the general description of the calculation and output in embodiment 1, the gain effect of this embodiment is that it provides a novel hybrid control architecture; the formula The non-obvious improvement of this formula is that:

[0183] Integration of feedback and feedforward: it not only contains feedback control responding to past errors , but also contains feedforward control responding to future predictions ;

[0184] Integration of control and strategy: it directly converts the macro-strategic decision of the upper layer S7 into gain adjustment of the lower layer S8 control law through ;

[0185] Multiplicative integration: using multiplication, not addition, to integrate the three terms, so that and can dynamically scale the strength of the basic control; this architecture has stronger nonlinearity and better decoupling characteristics than simple three-term addition, making the system achieve reactivity , foresight and strategy at the same time, which is a major improvement over traditional control laws.

[0186] Embodiment 8:

[0187] Please refer to Figure 2 A water quality guarantee management control system for a water dispenser, comprising:

[0188] A parameter acquisition module for acquiring multi-source service parameters including current water quality parameters, average water pump working pressure, and cumulative filter running time;

[0189] A water quality guarantee management module, comprising:

[0190] A prediction processing unit for obtaining short-term water quality prediction values and sudden pollution indexes through a preset deep learning time series prediction model based on the multi-source service parameters;

[0191] A health calculation unit for calculating a device health index based on a preset service life and a preset pressure by combining the cumulative filter running time and the average water pump working pressure;

[0192] A deviation determination unit for determining a composite water quality deviation by combining the current water quality parameters and the short-term water quality prediction values;

[0193] A risk calculation unit for determining a system risk index based on preset limits and weights by combining the composite water quality deviation, the device health index, and the sudden pollution indexes;

[0194] A grade division unit for dividing a risk grade based on the system risk index and the device health index and comparing preset thresholds;

[0195] A strategy generation unit for determining a healthy water quality trade-off factor in response to the risk grade;

[0196] A control output unit for determining a control target deviation based on the current water quality parameters and a preset target water quality, and calculating and outputting a final control output by combining the control target deviation, the healthy water quality trade-off factor, and the sudden pollution indexes.

[0197] The embodiment discloses a water quality guarantee management control system for a water dispenser, which is a physical carrier of the method described in embodiments 1-7; characterized in that the system comprises:

[0198] A parameter acquisition module for executing step S1 of embodiment 1; in this embodiment, the module is embodied as a set of sensors and a timer, including water quality sensors for obtaining current water quality parameters such as TDS and turbidity sensors, pressure sensors for obtaining average water pump working pressure , and a timer for recording cumulative filter running time ;

[0199] Water quality assurance management module, which aims to execute the core logic S2-S8 of embodiment 1; in this embodiment, the module is embodied as a central processor, such as a microcontroller MCU, an embedded system or a PLC, which is logically divided into the following functional units inside the processor:

[0200] Prediction processing unit: for executing S2; the unit is built-in with a preset deep learning time series prediction model, such as an RNN model file, which receives multi-source service parameters and calculates output short-term water quality prediction values and burst pollution index ;

[0201] Health calculation unit: for executing S3; the unit stores preset service life and preset stress , receives and , and executes the formula described in embodiment 2 to calculate the device health index ;

[0202] Deviation determination unit: for executing S4; the unit receives from the acquisition module, and from the prediction unit, and executes to determine the composite water quality deviation ;

[0203] Risk calculation unit: for executing S5; the unit stores preset limit value and preset weight , receives , and executes the formula described in embodiment 3 to determine the system risk index ;

[0204] Grade division unit: for executing S6; the unit stores preset threshold value , receives and , and executes the or logical judgment described in embodiment 4 to divide the risk grade;

[0205] Strategy generation unit: for executing S7; the unit receives the risk grade and determines the health water quality trade-off factor according to the mode switching logic of embodiment 5; if in a high-risk state, the unit will also call the formula of embodiment 6 for dynamic calculation;

[0206] Control output unit: for executing S8; the unit receives , and stores and gain , the formula described in embodiment 7 is executed , the final control output is calculated and output ;

[0207] The final control output is sent to the actuators of the water dispenser, such as the water pump driver and the purification unit controller, so as to change the running state of the system; the change of the state, the new and , is captured by the parameter acquisition module and fed back to the water quality guarantee management module, forming a complete adaptive closed-loop control;

[0208] The water quality guarantee management control system of the water dispenser described in the embodiment provides a physical implementation for the methods described in embodiments 1-7 through its clear modular architecture; each unit has a high degree of cohesion in function and clear logic; compared with the single-function controller in the prior art, the water quality guarantee management module of the system can execute the complete and advanced intelligent algorithm of prediction-evaluation-decision-control through its multiple collaborative units, so as to realize the adaptive closed-loop control of water quality safety, equipment health and running cost on the real water dispenser hardware, achieving the overall technical purpose of the application.

[0209] It should be noted that the above embodiments are only used to illustrate the technical solutions of the application and not to limit it. Although the application has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the application can be modified or replaced equivalently without departing from the spirit and scope of the technical solutions of the application.

Claims

1. A method for managing and controlling water quality in a water dispenser, characterized in that, include: S1 collects multi-source service parameters including current water quality parameters, average working pressure of water pumps, and cumulative operating time of filters; S2, based on multi-source service parameters, is processed by a pre-set deep learning time series prediction model to obtain short-term water quality prediction values ​​and sudden pollution index. S3, combining the cumulative operating time of the filter with the average working pressure of the water pump, and based on the preset lifespan and preset pressure, calculates the equipment health index; S4. Combine current water quality parameters with short-term water quality forecasts to determine the composite water quality deviation; S5, combining composite water quality deviation, equipment health index and sudden pollution index, determines the system risk index based on preset limits and weights; S6, based on the system risk index and equipment health index, and by comparing them with preset thresholds, the risk level is determined; S7, in response to risk level, determines healthy water quality trade-off factors; S8 determines the control target deviation based on the current water quality parameters and the preset target water quality, and calculates and outputs the final control output by combining the control target deviation, the healthy water quality trade-off factor and the sudden pollution index. The calculated equipment health index includes: The filter health parameters are determined based on the ratio of the filter's cumulative operating time to the preset rated reference life of the filter. The pump health parameters are determined based on the ratio of the pump's average operating pressure to the preset pump rated pressure. The equipment health index is obtained by weighted summing of the filter health items and the water pump health items; Determine the system risk index, including: Divide the composite water quality deviation by the preset water quality failure limit to obtain the water quality risk item; The loss item is used to calculate the equipment health index; The system risk index is obtained by weighting and summing the water quality risk item, the equipment health index loss item, and the sudden pollution index according to preset weights. The risk levels are divided into: When the system risk index is less than or equal to the preset normal operation threshold, the risk level is determined to be a safe state. If the system risk index is greater than the normal operation threshold but less than or equal to the preset failure boundary threshold, the risk level is determined to be high-risk. If the system risk index is greater than the failure boundary threshold, or the equipment health index is less than the preset health failure limit, the risk level is determined to be a failure state.

2. The water quality assurance management and control method for a water dispenser according to claim 1, characterized in that, Determine the trade-offs for healthy water quality, including: In response to the risk level being deemed safe, the healthy water quality tradeoff factor is determined to be a preset economic model value. In response to a high-risk status, a healthy water quality trade-off factor is calculated based on the equipment health index and the system risk index. In response to a risk level of failure, the healthy water quality trade-off factor is set to the preset strong purification mode value.

3. The water quality assurance management and control method for a water dispenser according to claim 2, characterized in that, Calculate and generate healthy water quality tradeoffs, including: The normalized risk factor is calculated based on the system risk index, the preset normal operation threshold, and the preset failure boundary threshold. The loss component of the equipment health index is calculated together with the normalized risk factor to obtain the healthy water quality trade-off factor.

4. The water quality assurance management and control method for a water dispenser according to claim 1, characterized in that, Calculate and output the final control output, including: Based on the deviation from the control target, a proportional feedback control term is generated; Based on the sudden pollution index, an AI feedforward gain term is generated; Based on healthy water quality trade-off factors, a healthy balance attenuation term is generated; By combining the proportional feedback control term, the AI ​​feedforward gain term, the health balance attenuation term, and the preset basic operating output, the final control output is calculated and output to the actuator.

5. A water quality assurance management and control system for a water dispenser, based on the water quality assurance management and control method for a water dispenser as described in any one of claims 1-4, characterized in that, include: The parameter acquisition module is used to collect multi-source service parameters, including current water quality parameters, average working pressure of water pumps, and cumulative operating time of filters. The water quality assurance management module includes: The prediction processing unit is used to obtain short-term water quality prediction values ​​and sudden pollution indexes based on multi-source service parameters and through a preset deep learning time series prediction model. The health calculation unit is used to combine the cumulative operating time of the filter with the average working pressure of the water pump, and calculate the equipment health index based on the preset lifespan and preset pressure. The deviation determination unit is used to determine the composite water quality deviation by combining the current water quality parameters with the short-term water quality forecast values. The risk calculation unit is used to determine the system risk index by combining composite water quality deviation, equipment health index and sudden pollution index, based on preset limits and weights. The risk level classification unit is used to classify risk levels based on the system risk index and the equipment health index, and by comparing them with preset thresholds. The strategy generation unit is used to determine healthy water quality trade-offs in response to risk levels. The control output unit is used to determine the control target deviation based on the current water quality parameters and the preset target water quality, and to calculate and output the final control output by combining the control target deviation, the healthy water quality trade-off factor and the sudden pollution index.

Images