Intelligent fish tank control system, control method and intelligent fish tank

Abstract

The invention relates to the technical field of fish tank intelligent control, in particular to an intelligent fish tank control system and method and an intelligent fish tank. The system collects fish tank water quality and environment parameters through a water quality sensing module to generate state vectors; the state analysis module constructs a parameter incidence matrix and quantifies mutual influence among parameters; the prediction modeling module is combined with the state vector and the incidence matrix to establish a water ecology dynamic evolution model and predict a multi-time-scale water quality change trend; the control decision module judges whether future water quality deviates from a healthy area or not according to the prediction result and a health threshold value and generates a control strategy; the execution control module receives the strategy, regulates and controls the water pump, the air pump and the lighting and feeding device, achieves intelligent regulation and maintenance of the water quality environment of the fish tank, achieves accurate prediction of the water quality change trend, converts a control mode from passive response to active prevention, remarkably improves the water quality stability and reduces extreme water quality events.

Description

Technical Field

[0001] This invention relates to the field of intelligent aquarium control technology, and particularly to an intelligent aquarium control system, control method, and intelligent aquarium. Background Technology

[0002] Traditional aquarium management relies mainly on manual, periodic water quality testing and adjustments, which is not only labor-intensive but also difficult to respond to changes in water quality in a timely manner. Existing automatic control systems, such as CN118859796B, disclose a water circulation control method and system based on water quality monitoring. This system can achieve automatic water quality optimization to a certain extent by monitoring water quality parameters and triggering water circulation control based on thresholds.

[0003] However, existing technologies have the following shortcomings: First, they only use simple threshold-triggered control, which lacks the ability to predict water quality change trends, causing the system to only passively respond to problems rather than actively prevent them; second, they treat each water quality parameter in isolation, ignoring the mutual influence and correlation between parameters, making it difficult to achieve overall coordinated control; third, the control strategy is singular, mainly achieving water quality control by adjusting water flow, which is difficult to adapt to complex and ever-changing aquatic ecological environments; fourth, they lack adaptive learning capabilities and cannot optimize control strategies based on operational experience.

[0004] With the development of artificial intelligence and Internet of Things technologies, there is an urgent need for an intelligent aquarium control system that can predict water quality changes and achieve proactive control in order to improve water quality stability, reduce energy consumption, and enhance the fishkeeping experience. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent aquarium control system, control method, and intelligent aquarium, aiming to solve the problems of passive response, isolated parameter processing, and single control strategy in the prior art.

[0006] This invention discloses an intelligent fish tank control system, comprising:

[0007] The water quality sensing module is used to collect water quality parameters and environmental parameters in the aquarium and generate a water quality state vector.

[0008] The state analysis module is communicatively connected to the water quality sensing module and is used to receive the water quality state vector and construct a parameter correlation matrix. The parameter correlation matrix is ​​used to quantify the mutual influence relationship between different water quality parameters.

[0009] The predictive modeling module is communicatively connected to the state analysis module and is used to establish a dynamic evolution model of the water ecology based on the water quality state vector and the parameter correlation matrix, predict the changing trends of water quality parameters at multiple time scales, and generate prediction results.

[0010] The control decision module, communicatively connected to the prediction modeling module, receives the prediction results, determines whether the future water quality status will deviate from the healthy zone based on a preset health status threshold, and generates a control strategy; and

[0011] The execution control module is communicatively connected to the control decision module and is used to receive the control strategy, control the working status of the water pump, air pump, lighting and feeding device, and adjust the water quality environment in the fish tank.

[0012] Preferably, the water quality sensing module includes:

[0013] A multi-parameter sensor array is used to collect parameters such as pH, dissolved oxygen, ammonia nitrogen, nitrite, nitrate, temperature, and turbidity.

[0014] A signal processing unit, connected to the multi-parameter sensor array, is used for filtering, calibrating, and detecting anomalies in the raw sensor data; and

[0015] The state vector construction unit, connected to the signal processing unit, is used to integrate the processed multi-parameter data into a unified water quality state vector and add timestamp information.

[0016] Preferably, the state analysis module includes:

[0017] The correlation analysis unit is used to calculate the mutual influence coefficients between different water quality parameters and construct the parameter correlation matrix.

[0018] A health status assessment unit is used to calculate the health index of the current water quality status based on the ideal range of water quality parameters; and

[0019] The trend analysis unit is used to identify the changing trends and abnormal patterns of water quality parameters.

[0020] Preferably, the predictive modeling module includes:

[0021] Multi-timescale state evolution unit is used to establish the evolution equation of water quality parameters over time, and the parameters are divided into rapidly changing parameters, medium-speed changing parameters and slowly changing parameters;

[0022] The equilibrium analysis unit is used to identify the equilibrium and critical state characteristics of a system; and

[0023] The prediction engine unit is used to combine historical data and model parameters to generate short-term, medium-term, and long-term water quality status predictions.

[0024] Preferably, the control decision module includes:

[0025] The prediction and assessment unit is used to assess whether the prediction results indicate that the water quality status will deviate from the healthy zone.

[0026] A multi-objective optimization unit is used to balance multiple objectives such as water quality stability, energy efficiency, intervention minimization, and biological comfort; and

[0027] The intervention strategy unit is used to determine the optimal timing and intensity of intervention and to generate specific control strategies.

[0028] Preferably, the execution control module includes:

[0029] The instruction conversion unit is used to convert abstract control strategies into specific actuator instructions;

[0030] The execution scheduling unit is used to arrange the working sequence of each actuator; and

[0031] The feedback monitoring unit is used to monitor the effectiveness of the control and adjust the execution plan when necessary.

[0032] An intelligent aquarium control method, employing the aforementioned system, includes:

[0033] Collect water quality and environmental parameters in the aquarium to generate a water quality state vector;

[0034] Construct a parameter correlation matrix, which is used to quantify the mutual influence relationship between different water quality parameters;

[0035] Based on the water quality state vector and the parameter correlation matrix, a water ecological dynamic evolution model is established to predict the changing trends of water quality parameters at multiple time scales and generate prediction results.

[0036] Upon receiving the prediction results, and based on a preset health status threshold, determine whether the future water quality will deviate from the healthy range, and generate a control strategy; and

[0037] According to the control strategy, the working status of the water pump, air pump, lighting and feeding device are controlled to adjust the water quality environment in the fish tank.

[0038] Preferably, the prediction of water quality parameter change trends across multiple time scales includes:

[0039] Establish the evolution equation of the state vector over time: S(t+Δt)=F(S(t),C(t),Δt), where S(t) is the current state vector, C(t) is the control vector, F is the state transition function, and Δt is the time step.

[0040] Based on the characteristics of parameter changes, parameters are divided into rapidly changing parameters, medium-changing parameters, and slowly changing parameters; and by combining short-term historical data and long-term patterns, prediction results for 1 hour, 6 hours, 24 hours, and 72 hours are generated.

[0041] Preferably, the generation control strategy includes:

[0042] Multiple control objectives are set, including water quality stability, energy efficiency, minimal intervention, and biological comfort;

[0043] Construct the objective function ,

[0044] in, Represents the overall objective function. Indicates water quality stability target, Indicates energy efficiency targets. This indicates the goal of minimizing intervention. Represents the biological comfort goals, with the weights for each goal;

[0045] Under the constraints of actuator physical limitations, parameter change rate limitations, and energy consumption upper limit, the optimal control vector is solved; and a gradual intervention strategy is designed, which first makes slight adjustments and then decides whether to strengthen the intervention based on the feedback results.

[0046] Intelligent fish tanks include:

[0047] The fish tank itself;

[0048] A multi-parameter sensor array is installed on the aquarium body to collect parameters such as pH value, dissolved oxygen, ammonia nitrogen, nitrite, nitrate, temperature and turbidity;

[0049] The actuator system installed on the aquarium body includes a variable frequency water pump, a gas exchange device, an LED lighting system, and an automatic feeding device; and

[0050] A control system is connected to the multi-parameter sensor array and the actuator system, and the control system includes the intelligent aquarium control system.

[0051] The beneficial effects of this invention include:

[0052] 1. To achieve accurate prediction of water quality change trends, shift the control mode from passive response to proactive prevention, significantly improve water quality stability, and reduce the occurrence of extreme water quality events;

[0053] 2. By establishing a parameter correlation matrix, the mutual influence relationship between different water quality parameters can be quantified, achieving overall coordinated control and avoiding chain reactions caused by single parameter adjustment;

[0054] 3. A multi-objective optimization control strategy is adopted to ensure water quality stability while taking into account energy efficiency and biological comfort, thereby significantly reducing system energy consumption and maintenance costs.

[0055] 4. Based on a dynamic evolution model with multiple time scales, it can simultaneously handle rapidly changing and slowly changing water quality parameters, achieve full-time domain control, and improve the system's adaptability and robustness. Attached Figure Description

[0056] Figure 1 This is a structural block diagram of the intelligent fish tank control system provided in an embodiment of the present invention;

[0057] Figure 2 This is a structural block diagram of the water quality sensing module provided in an embodiment of the present invention;

[0058] Figure 3 This is a structural block diagram of the state analysis module provided in an embodiment of the present invention;

[0059] Figure 4 This is a structural block diagram of the prediction modeling module provided in an embodiment of the present invention;

[0060] Figure 5 This is a structural block diagram of the control decision module provided in an embodiment of the present invention;

[0061] Figure 6 This is a structural block diagram of the execution control module provided in an embodiment of the present invention;

[0062] Figure 7 A flowchart of an intelligent fish tank control method provided in an embodiment of the present invention. Detailed Implementation

[0063] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0064] Reference Figure 1 The intelligent aquarium control system provided in this embodiment of the invention includes: a water quality sensing module 10, a status analysis module 20, a prediction modeling module 30, a control decision module 40, and an execution control module 50.

[0065] The water quality sensing module 10 is used to collect water quality parameters and environmental parameters in the fish tank and generate a water quality state vector.

[0066] Reference Figure 2 In a preferred embodiment of the present invention, the water quality sensing module 10 includes a multi-parameter sensor array 11, a signal processing unit 12, and a state vector construction unit 13.

[0067] The multi-parameter sensor array 11 is used to collect parameters such as pH, dissolved oxygen, ammonia nitrogen, nitrite, nitrate, temperature, and turbidity. Preferably, the acquisition accuracy for pH is ±0.1, for dissolved oxygen it is ±0.1 mg / L, and for temperature it is ±0.2℃. For critical parameters such as pH and dissolved oxygen, the acquisition frequency can be set to a higher value, for example, once every 30 seconds; while for parameters that change slowly, such as nitrate, the acquisition frequency can be set to a lower value, for example, once every 30 minutes.

[0068] The signal processing unit 12 is connected to the multi-parameter sensor array 11 and is used to filter, calibrate, and detect anomalies in the raw sensor data. In one embodiment, the signal processing unit 12 uses a sliding window mean filtering algorithm to remove short-term noise. The window size is dynamically adjusted according to the parameter change characteristics; for example, the window size can be set to 5 for pH and 3 for temperature. Simultaneously, the signal processing unit 12 also detects data anomalies. If a parameter value changes abruptly beyond a preset threshold (e.g., pH changes by more than 0.5 within one minute), it is marked as abnormal data and not included in subsequent processing.

[0069] The state vector construction unit 13 is connected to the signal processing unit 12 and is used to integrate the processed multi-parameter data into a unified water quality state vector and add timestamp information. The water quality state vector can be represented as:

[0070] ,

[0071] in, Let be the water quality state vector at time t. This represents the pH value at time t (dimensionless, range 0-14). This represents the dissolved oxygen concentration (mg / L) at time t. This represents the ammonia nitrogen concentration (mg / L) at time t. This represents the nitrite concentration (mg / L) at time t. This represents the nitrate concentration (mg / L) at time t. This represents the water temperature (°C) at time t. The turbidity at time t (NTU, i.e., turbidity unit) The organic load index (dimensionless value from 0 to 100) represents the organic load index at time t. The nitrification efficiency index (dimensionless value of O-100) represents the nitrification efficiency index at time t. Indicates time (seconds).

[0072] It should be noted that the organic matter load index and nitrification efficiency index These are derived indicators calculated based on basic water quality parameters. The organic matter load index can be calculated by combining turbidity, pH change rate, and dissolved oxygen consumption rate; the nitrification efficiency index reflects the efficiency of ammonia nitrogen to nitrite and nitrite to nitrate conversion, and can be calculated by continuously monitoring the change rate of these three parameters.

[0073] The state analysis module 20 is communicatively connected to the water quality sensing module 10 and is used to receive the water quality state vector and construct a parameter correlation matrix. The parameter correlation matrix is ​​used to quantify the mutual influence relationship between different water quality parameters.

[0074] Reference Figure 3 In one embodiment of the present invention, the status analysis module 20 includes a correlation analysis unit 21, a health status assessment unit 22, and a trend analysis unit 23.

[0075] The correlation analysis unit 21 is used to calculate the mutual influence coefficients between different water quality parameters and construct the parameter correlation matrix. The parameter correlation matrix R can be represented as an n×n matrix (n is the number of water quality parameters), where the elements... This represents the influence coefficient of parameter i. For example, This represents the coefficient of influence of dissolved oxygen on pH value.

[0076] The correlation analysis unit 21 determines the influence coefficient by analyzing the correlation of parameter changes in historical data. Specifically, the following methods can be used:

[0077] ,

[0078] in, Indicates parameters For parameters The influence coefficient (dimensionless) Indicates parameters exist The value at time, Indicates parameters exist The value at time, Indicates parameters The average value during the observation period, Indicates the time point sequence number. Indicates the total number of data points. Indicates time delay (unit: minutes or hours).

[0079] Preferably, It can be adjusted according to the rate of interaction between different parameters, such as the effect of dissolved oxygen on pH. It can be set to 10 minutes; as for the effect of organic load on nitrate, It can be set to 24 hours. The health status assessment unit 22 is used to calculate the health index of the current water quality status based on the ideal range of water quality parameters. For each parameter... Define its ideal range and optimal value Health Index It can be calculated as:

[0080] ,

[0081] in, This represents the health index (a dimensionless value from 0 to 1, with 1 representing the healthiest). Indicates the total number of parameters. For parameters The weighting coefficients (dimensionless, and satisfying) ), For parameters The deviation (dimensionless value of 0-1) is calculated as follows:

[0082] ,

[0083] in, This indicates the deviation of parameter i. This represents the current value of parameter i. This represents the optimal value of parameter i. This represents the lower limit of the ideal range of parameter i. This represents the upper limit of the ideal range of parameter i. When the parameter value equals the optimal value, the deviation is 0; when the parameter value reaches or exceeds the boundary of the ideal range, the deviation is 1.

[0084] Preferably, for tropical fish farming, the ideal ranges for each parameter can be set as follows: pH value [6.5, 7.5], with an optimal value of 7.0; dissolved oxygen [5.0, 0.05 mg / L]; nitrate [0, 40] mg / L, with an optimal value of 10 mg / L; and temperature [24, 28] °C, with an optimal value of 26 °C.

[0085] The trend analysis unit 23 is used to identify the changing trends and abnormal patterns of water quality parameters. By calculating the first derivative (rate of change) and second derivative (acceleration of change) of the parameters, the trend analysis unit 23 can identify whether the parameters are rising, falling, or stable, and whether the change is accelerating or decelerating. However, the trend analysis unit 23 cannot identify specific abnormal patterns, such as periodic fluctuations or abrupt changes.

[0086] The predictive modeling module 30 is connected to the state analysis module 20 to establish a dynamic evolution model of the water ecosystem based on the water quality state vector and parameter correlation matrix, predict the changing trends of water quality parameters at multiple time scales, and generate prediction results.

[0087] Reference Figure 4 In one embodiment of the present invention, the prediction modeling module 30 includes a multi-timescale state evolution unit 31, an equilibrium state analysis unit 32, and a prediction engine unit 33.

[0088] The multi-timescale state evolution unit 31 is used to establish the evolution equations of water quality parameters over time, classifying the parameters into rapidly changing parameters, moderately changing parameters, and slowly changing parameters. The state evolution equations can be expressed as:

[0089] The multi-timescale state evolution unit 31 is used to establish the evolution equations of water quality parameters over time, classifying the parameters into rapidly changing parameters, moderately changing parameters, and slowly changing parameters. The state evolution equations can be expressed as:

[0090] ,

[0091] in, Let be the state vector at time t. Let be the state vector at time t+Δt. This is the control vector at time t (containing control variables such as water flow rate, aeration rate, and light intensity). Δt is the state transition function, and Δt is the time step (in minutes or hours).

[0092] Considering the significant differences in the rate of change of different parameters, the multi-timescale state evolution unit 31 divides the parameters into three categories: rapidly changing parameters (such as dissolved oxygen and pH): the change characteristic time is on the order of minutes, corresponding to Δt=1 minute; medium-speed changing parameters (such as ammonia nitrogen and temperature): the change characteristic time is on the order of hours, corresponding to Δt=1 hour; and slowly changing parameters (such as nitrate and biofilm formation): the change characteristic time is on the order of days, corresponding to Δt=1 day.

[0093] The state transition function F can be expressed as:

[0094] ,

[0095] Where A is the parameter natural rate of change matrix (n×n matrix, where n is the number of parameters), reflecting the natural change trend of parameters without external intervention; B is the control influence matrix (n×m matrix, where m is the number of control variables), representing the direct influence of control variables on parameters; D is the parameter interaction influence matrix (n×n matrix), representing the strength of interaction between parameters; and R is the parameter correlation matrix (n×n matrix), which is the parameter correlation matrix generated by the state analysis module. The Hadamard product (element-by-element multiplication) is used to calculate the nonlinear interaction effects between parameters; E(t) is the external disturbance term (an n-dimensional vector) representing the random influence of external environmental factors. These matrix parameters are obtained through training with historical data. The equilibrium analysis unit 32 is used to identify the equilibrium and critical state characteristics of the system.

[0096] An equilibrium state refers to a state that satisfies the condition ||S(t+Δt)-S(t)||<ε, where ||S(t+Δt)-S(t)|| represents the Euclidean norm change of the state vector in time Δt, and ε is a preset threshold (usually set to 5% of the ideal range width of each element of the state vector).

[0097] The equilibrium analysis unit 32 also defines a system stability index λ, which is used to quantify the system's resistance to external disturbances:

[0098] ,

[0099] in, Indicators representing system stability (units depend on the units of the parameters), This represents the i-th component of the state transition function. This represents the i-th component of the state vector. Representation function For variables The partial derivatives, This means taking the minimum value among all values ​​of i. The higher the value, the higher the system stability. In addition, the equilibrium analysis unit 32 can identify critical state characteristics, such as precursor signals like a sudden slowdown in the rate of parameter change or an increase in fluctuation amplitude, providing early warning for the system.

[0100] Prediction engine unit 33 combines historical data and model parameters to generate short-term, medium-term, and long-term water quality status predictions. Prediction engine unit 33 employs an iterative method for prediction.

[0101] 1. For short-term forecasts (1-6 hours), use hourly time steps and a complete state transition function;

[0102] 2. For medium-term forecasts (6–24 hours), a larger time step is used to simplify some rapidly changing processes;

[0103] 3. For long-term forecasts (24–72 hours), focus on the trends of slowly changing parameters.

[0104] The forecast results include predicted values ​​and forecast intervals. Forecast interval The calculations, which take into account model errors and external disturbance uncertainties, provide a risk assessment basis for control decisions. This represents the lower bound of the state vector at time t+Δt. It represents the upper bound of the state vector at time t+Δt.

[0105] The control decision module 40 is communicatively connected to the prediction modeling module 30. It is used to receive prediction results, determine whether the future water quality status will deviate from the healthy zone based on the preset health status threshold, and generate control strategies.

[0106] Reference Figure 5 In one embodiment of the present invention, the control decision module 40 includes a prediction and evaluation unit 41, a multi-objective optimization unit 42, and an intervention strategy unit 43.

[0107] The prediction and assessment unit 41 is used to assess whether the prediction results indicate that the water quality status will deviate from the healthy zone. Specifically, the prediction and assessment unit 41 calculates the health index of the predicted status. and with preset threshold (Usually set to 0.8) Comparison. If If this is the case, it indicates that the future water quality will deviate from the healthy range, requiring intervention. Among these, The health index (a dimensionless value of 0-1) represents the predicted state. This represents the health index threshold (a dimensionless value between 0 and 1).

[0108] The multi-objective optimization unit 42 is used to balance multiple objectives such as water quality stability, energy efficiency, intervention minimization, and biological comfort. The multi-objective optimization unit 42 constructs the objective function:

[0109] ,

[0110] in, This represents the overall objective function (dimensionless). This represents the target for water quality stability (the expected value of the health index, a dimensionless value between 0 and 1). This represents the energy efficiency target (a negative value for energy consumption, in units of -W or -kWh). This represents the intervention minimization target (the negative norm of the change in the control quantity, the unit of which depends on the unit of the control quantity). This represents the biological comfort target (an environmental parameter stability index, a dimensionless value). The weights of each objective (dimensionless) satisfy the following conditions: .

[0111] In a preferred embodiment, the weights of each target can be set as follows: , , , This reflects that water quality stability is the primary goal, while other factors are also taken into account.

[0112] The multi-objective optimization unit 42 solves for the optimal control vector under the following constraints. :

[0113] Actuator physical limitations: such as water pump flow rate range [0.5, 8] L / min;

[0114] Parameter change rate limit: For example, the pH value should not change by more than 0.3 per hour;

[0115] Energy consumption limit: such as total power not exceeding 30W;

[0116] Intervention strategy unit 43 is used to determine the optimal timing and intensity of intervention, and to generate specific control strategies. Intervention strategy unit 43 divides control into three levels:

[0117] Intervention strategy unit 43 is used to determine the optimal timing and intensity of intervention, and to generate specific control strategies. Intervention strategy unit 43 divides control into three levels:

[0118] Preventive control: Taking fine-tuning measures in advance to address anticipated potential problems;

[0119] Corrective control: Appropriate intervention for problems that have deviated but are not yet severe.

[0120] Emergency control: In response to a rapidly deteriorating situation, take strong intervention measures;

[0121] For preventative control, intervention strategy unit 43 designs a gradual intervention strategy, starting with slight adjustments (such as increasing water flow by 10%), observing feedback, and then deciding whether to strengthen the intervention. This approach minimizes unnecessary intervention, reduces energy consumption, and minimizes stress on the fish.

[0122] The execution control module 50 is communicatively connected to the control decision module 40, and is used to receive control strategies, control the working status of water pumps, air pumps, lighting and feeding devices, and adjust the water quality environment in the aquarium.

[0123] Reference Figure 6 In one embodiment of the present invention, the execution control module 50 includes an instruction conversion unit 51, an execution scheduling unit 52, and a feedback monitoring unit 53.

[0124] The instruction conversion unit 51 is used to convert abstract control strategies into specific actuator instructions. For example, increasing water flow by 20% is converted into a specific water pump speed adjustment instruction; reducing light intensity by 15% is converted into an LED drive current adjustment instruction.

[0125] The execution scheduling unit 52 is used to arrange the working sequence of each actuator. Some control operations need to be executed in a specific order, such as increasing the aeration rate before increasing the feeding rate to prevent a sudden drop in dissolved oxygen. The execution scheduling unit 52 determines the optimal execution order based on the priority and dependencies of the operations.

[0126] The feedback monitoring unit 53 is used to monitor the effectiveness of the control execution and adjust the execution plan when necessary. The feedback monitoring unit 53 compares the parameter changes with the expected results in real time. If the deviation exceeds the threshold (for example, the expected pH value increases by 0.1, but the actual increase is only 0.02), it will trigger an adjustment to the execution plan, such as increasing the intensity of the intervention or extending the intervention time.

[0127] Reference Figure 7 This invention also provides an intelligent fish tank control method, comprising the following steps:

[0128] Step S701: Collect water quality parameters and environmental parameters in the fish tank to generate a water quality state vector.

[0129] In this step, parameters such as pH, dissolved oxygen, ammonia nitrogen, nitrite, nitrate, temperature and turbidity are collected by a multi-parameter sensor array. The raw sensor data is filtered, calibrated and anomaly detected. Then, the processed multi-parameter data is integrated into a unified water quality state vector and timestamp information is added.

[0130] Step S702: Construct a parameter correlation matrix, which is used to quantify the mutual influence relationship between different water quality parameters.

[0131] In this step, the correlation between parameter changes in historical data is analyzed, the mutual influence coefficients between different water quality parameters are calculated, and a parameter correlation matrix is ​​constructed.

[0132] Step S703: Based on the water quality state vector and the parameter correlation matrix, establish a water ecological dynamic evolution model, predict the changing trends of water quality parameters at multiple time scales, and generate prediction results.

[0133] In this step, the evolution equation of the state vector over time is established as S(t+Δt)=F(S(t),C(t),Δt), where S(t) is the current state vector, C(t) is the control vector, and F is the state transition function. Based on the parameter change characteristics, the parameters are divided into rapidly changing parameters, medium-changing parameters, and slowly changing parameters, corresponding to minute-level, hour-level, and day-level change characteristic times, respectively. Combining short-term historical data and long-term models, prediction results for 1 hour, 6 hours, 24 hours, and 72 hours are generated.

[0134] Step S704: Receive the prediction result, determine whether the future water quality status will deviate from the healthy zone based on the preset health status threshold, and generate a control strategy.

[0135] In this step, the predicted health index is evaluated and compared with a preset threshold. If the predicted health index is lower than the threshold, it is determined that the future water quality will deviate from the healthy zone, requiring intervention. Multiple control objectives are set, including water quality stability, energy efficiency, intervention minimization, and biological comfort, constructing the objective function J = w1·J1 + w2·J2 + w3·J3 + w4·J4, where w1 represents the weight of each objective. Under constraints such as actuator physical limitations, parameter change rate limits, and energy consumption upper limits, the optimal control vector is solved.

[0136] Design a gradual intervention strategy, starting with minor adjustments and then deciding whether to increase the intervention based on feedback. This approach minimizes unnecessary intervention, reduces energy consumption, and minimizes stress on the fish.

[0137] Step S705: According to the control strategy, control the working status of the water pump, air pump, lighting and feeding device to adjust the water quality environment in the fish tank.

[0138] In this step, the abstract control strategy is converted into specific actuator instructions, the working sequence of each actuator is arranged, the control execution effect is monitored, and the execution plan is adjusted as necessary.

[0139] This invention also provides an intelligent fish tank, including a fish tank body, a multi-parameter sensor array disposed on the fish tank body, an actuator system disposed on the fish tank body, and a control system.

[0140] A multi-parameter sensor array is used to collect parameters such as pH, dissolved oxygen, ammonia nitrogen, nitrite, nitrate, temperature, and turbidity. In a preferred embodiment, the sensors are waterproof and transmit data wirelessly, allowing for flexible placement in different locations within the aquarium, avoiding wiring difficulties and aesthetic issues.

[0141] The actuator system includes a variable frequency water pump, a gas exchange device, an LED lighting system, and an automatic feeding device. The variable frequency water pump can achieve a flow rate range of 0.5–8 L / min with an accuracy of ±0.1 L / min; the gas exchange device includes a microporous aerator and a CO2 controller; the LED lighting system supports color temperature adjustment from 6500K to 9000K and brightness adjustment from 0% to 100%; the automatic feeding device can release precisely measured feed at specific times according to a set schedule.

[0142] The control system connects to a multi-parameter sensor array and actuator system, including the aforementioned intelligent aquarium control system. The control system hardware employs a 32-bit microcontroller (such as the STM32F4 series) with a 180MHz clock speed, equipped with 8MB SRAM and 32MB flash memory to meet complex computing and data storage requirements. The system achieves wireless communication via Wi-Fi / Bluetooth modules, supporting remote monitoring and control. The user interface includes a 5-inch color touchscreen display that intuitively shows water quality status and predictive information, supporting parameter settings and control mode selection.

[0143] Example 1: 60L Intelligent Ecological Fish Tank for Ornamental Fish

[0144] This embodiment provides a 60L intelligent ecological aquarium for ornamental fish, with the following configuration:

[0145] 60×30×35cm glass fish tank (approximately 60L of water);

[0146] Main controller: STM32F429 chip, 180MHz clock speed, 256KB RAM;

[0147] Integrated water quality sensor (pH, dissolved oxygen, temperature);

[0148] Ion-selective electrode (ammonia nitrogen, nitrite);

[0149] Variable frequency water pump (0.5~5L / min);

[0150] Dual-channel peristaltic pump (pH adjustment);

[0151] Adjustable spectrum LED lighting (18W);

[0152] The workflow is as follows:

[0153] (1) Learning adaptation stage:

[0154] After the system is started, it enters a 7-day learning and adaptation period, during which water quality data is collected at high frequency (pH and dissolved oxygen every 5 minutes, other parameters every 30 minutes), the daily change patterns and interrelationships of each parameter are recorded, and the initial parameter correlation matrix and dynamic model parameters are constructed.

[0155] (2) Daily monitoring mode:

[0156] The system enters routine monitoring mode, continuously monitoring water quality parameters and performing a comprehensive predictive analysis every 6 hours, displaying real-time water quality status and predicted trends for the next 24 hours. The system is in automatic control mode, automatically adjusting environmental parameters based on the prediction results.

[0157] (3) Examples of prevention and intervention:

[0158] The system detected a slight increase in ammonia nitrogen (0.2 mg / L) at 10:00 AM on Tuesday. Predictive analysis indicated that without intervention, ammonia nitrogen would reach the warning level (0.5 mg / L) after 15 hours. The system evaluated several intervention options and selected the optimal one: increasing the water flow rate by 20% (from 2 L / min to 2.4 L / min) for 2 hours; and reducing the next feeding amount by 15%. After the intervention, ammonia nitrogen changes were continuously monitored. After 24 hours, ammonia nitrogen stabilized at 0.25 mg / L, not reaching the warning level. The system recorded this successful intervention and updated the model parameters.

[0159] (4) Example of exception handling:

[0160] At 20:00 on Thursday, the pH level suddenly dropped (from 7.2 to 6.8). The system detected this as an abnormal change and triggered an alert. An emergency analysis process was initiated, revealing a simultaneous drop in dissolved oxygen. This was diagnosed as a possible surge in bioburden or equipment malfunction. Emergency intervention was implemented: aeration was increased by 30%, and 3 ml of alkaline buffer solution was added. Users were notified of the potential problem, and a check of the filtration system was recommended. Intensive monitoring (pH every 2 minutes) was conducted until parameters returned to a safe range.

[0161] Example 2: 120L Professional Smart Fish Tank

[0162] This embodiment provides a 120L smart fish tank suitable for professional aquaculture, which has the following differences in configuration compared to Embodiment 1:

[0163] Expand the sensor network: Add nitrate, phosphate, and redox potential (ORP) sensors;

[0164] Enhanced computing platform: dual-core processor, 512KB RAM, 128MB storage capacity;

[0165] High-precision multi-channel peristaltic pump system (4 channels);

[0166] A zoned water flow control system creates different water flow environments;

[0167] Automatic water change system, supporting timed and triggered water changes;

[0168] Special features:

[0169] Multi-species ecological balance management: Supports the simultaneous farming of fish species with different needs;

[0170] Breeding cycle management: Provides specialized settings for breeding environment parameters;

[0171] Advanced early warning system: Predicts potential problems 48-72 hours in advance;

[0172] Water quality stability analysis: Generate a long-term water quality stability report;

[0173] This embodiment is applicable to professional fish breeding and aquaculture scenarios with high water quality requirements, and can more accurately manage complex aquatic ecosystems.

[0174] The intelligent aquarium control system and intelligent aquarium of the present invention, through the establishment of a multi-parameter correlation aquatic ecological state characterization system, an aquatic ecological dynamic evolution prediction model, and a prediction-based multi-objective optimization control strategy, achieve a technological leap from passive response to active prediction, and have the following significant technical effects:

[0175] 1. Significantly improved water quality stability: The fluctuation range of key water quality parameters (pH, ammonia nitrogen, nitrite) is reduced by 70%, the proportion of time that water quality parameters are maintained within the ideal range is increased to 93%, and the number of harmful parameter mutation events is reduced by 85%. Compared with traditional single-parameter threshold control systems, this invention establishes a parameter correlation network to quantify the mutual influence between parameters, thereby achieving overall coordinated control and fundamentally improving system stability.

[0176] 2. Significantly improved energy efficiency: Pump operating time is reduced by 55% while maintaining better water quality; total energy consumption is reduced by 45%, especially during low-activity periods at night; water treatment chemical usage is reduced by 70%, and water change frequency is reduced by 60%. The predictive control strategy of this invention can detect potential problems in advance and prevent them from escalating through small-scale, timely interventions, thereby significantly reducing energy and resource consumption.

[0177] 3. Improved fish health: Disease incidence is reduced by 65%, and average lifespan is extended by 30%, with common water quality-related diseases almost eliminated. By maintaining a more stable water quality environment and reducing stress factors, this invention significantly improves the living conditions of fish.

[0178] 4. Comprehensive Improvement in User Experience: System operation complexity is reduced by 75%, maintenance time is reduced by 65%, and most maintenance work is completed automatically by the system; water quality prediction and health assessment are provided, allowing users to understand the system status in advance; the accuracy rate of abnormal situation early warning reaches 92%, almost eliminating sudden fish deaths. The predictive control and intuitive status display of this invention transform complex water quality management into a simple monitoring process, greatly reducing the professional knowledge requirements and daily maintenance burden for users.

[0179] The above technical effects have been verified through actual testing, and they have performed excellently in various aquarium environments, especially in professional aquaculture scenarios with strict water quality requirements.

[0180] The embodiments described above are merely illustrative of specific implementations of the present invention, and while the descriptions are detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention.

Claims

1. An intelligent fish tank control system, characterized in that, include: The water quality sensing module is used to collect water quality parameters and environmental parameters in the aquarium and generate a water quality state vector. The state analysis module is communicatively connected to the water quality sensing module and is used to receive the water quality state vector and construct a parameter correlation matrix. The parameter correlation matrix is ​​used to quantify the mutual influence relationship between different water quality parameters. The predictive modeling module is communicatively connected to the state analysis module and is used to establish a dynamic evolution model of the water ecology based on the water quality state vector and the parameter correlation matrix, predict the changing trends of water quality parameters at multiple time scales, and generate prediction results. The control decision module is communicatively connected to the prediction modeling module. It is used to receive the prediction results, determine whether the future water quality status will deviate from the healthy zone based on the preset health status threshold, and generate a control strategy. And an execution control module, which is communicatively connected to the control decision module, is used to receive the control strategy, control the working status of the water pump, air pump, lighting and feeding device, and adjust the water quality environment in the fish tank.

2. The intelligent fish tank control system according to claim 1, characterized in that, The water quality sensing module includes: A multi-parameter sensor array is used to collect parameters such as pH, dissolved oxygen, ammonia nitrogen, nitrite, nitrate, temperature, and turbidity. A signal processing unit, connected to the multi-parameter sensor array, is used to filter, calibrate, and detect anomalies in the raw sensor data; and a state vector construction unit, connected to the signal processing unit, is used to integrate the processed multi-parameter data into a unified water quality state vector and add timestamp information.

3. The intelligent fish tank control system according to claim 1, characterized in that, The status analysis module includes: The correlation analysis unit is used to calculate the mutual influence coefficients between different water quality parameters and construct the parameter correlation matrix. The health status assessment unit is used to calculate the current water quality health index based on the ideal range of water quality parameters; and the trend analysis unit is used to identify the changing trends and abnormal patterns of water quality parameters.

4. The intelligent fish tank control system according to claim 1, characterized in that, The predictive modeling module includes: Multi-timescale state evolution unit is used to establish the evolution equation of water quality parameters over time, and the parameters are divided into rapidly changing parameters, medium-speed changing parameters and slowly changing parameters; The equilibrium analysis unit is used to identify the equilibrium and critical state characteristics of the system; and the prediction engine unit is used to combine historical data and model parameters to generate short-term, medium-term and long-term water quality state prediction results.

5. The intelligent fish tank control system according to claim 1, characterized in that, The control decision module includes: The prediction and assessment unit is used to assess whether the prediction results indicate that the water quality status will deviate from the healthy zone. The multi-objective optimization unit is used to balance multiple objectives such as water quality stability, energy efficiency, intervention minimization and biological comfort; and the intervention strategy unit is used to determine the optimal intervention timing and intensity and generate specific control strategies.

6. The intelligent fish tank control system according to claim 1, characterized in that, The execution control module includes: The instruction conversion unit is used to convert abstract control strategies into specific actuator instructions; The execution scheduling unit is used to arrange the working sequence of each actuator; and the feedback monitoring unit is used to monitor and control the execution effect and adjust the execution plan when necessary.

7. An intelligent fish tank control method, employing the system described in any one of claims 1-6, characterized in that, include: Collect water quality and environmental parameters in the aquarium to generate a water quality state vector; Construct a parameter correlation matrix, which is used to quantify the mutual influence relationship between different water quality parameters; Based on the water quality state vector and the parameter correlation matrix, a water ecological dynamic evolution model is established to predict the changing trends of water quality parameters at multiple time scales and generate prediction results. Upon receiving the prediction results, and based on a preset health status threshold, determine whether the future water quality status will deviate from the healthy zone, and generate a control strategy. And according to the control strategy, control the working status of the water pump, air pump, lighting and feeding device, and adjust the water quality environment in the fish tank.

8. The intelligent fish tank control method according to claim 7, characterized in that, The predicted trends of water quality parameters across multiple time scales include: Establish the evolution equation of the state vector over time: S(t+Δt)=F(S(t),C(t),Δt), where S(t) is the current state vector, C(t) is the control vector, F is the state transition function, and Δt is the time step. Based on the characteristics of parameter changes, parameters are divided into rapidly changing parameters, medium-changing parameters, and slowly changing parameters; and by combining short-term historical data and long-term patterns, prediction results for 1 hour, 6 hours, 24 hours, and 72 hours are generated.

9. The intelligent fish tank control method according to claim 7, characterized in that, The generation control strategy includes: Multiple control objectives are set, including water quality stability, energy efficiency, minimal intervention, and biological comfort; Construct the objective function , in, Represents the overall objective function. Indicates water quality stability target, Indicates energy efficiency targets. This indicates the goal of minimizing intervention. Represents the biological comfort goals, with the weights for each goal; Under the constraints of actuator physical limitations, parameter change rate limitations, and energy consumption upper limit, the optimal control vector is solved; and a gradual intervention strategy is designed, which first makes slight adjustments and then decides whether to strengthen the intervention based on the feedback results.

10. An intelligent fish tank, characterized in that, include: The fish tank itself; A multi-parameter sensor array is installed on the aquarium body to collect parameters such as pH value, dissolved oxygen, ammonia nitrogen, nitrite, nitrate, temperature and turbidity. An actuator system installed on the aquarium body includes a variable frequency water pump, a gas exchange device, an LED lighting system, and an automatic feeding device; and a control system connected to the multi-parameter sensor array and the actuator system, wherein the control system includes the intelligent aquarium control system as described in any one of claims 1-6.

Images