A breeding fermentation bed temperature regulating system

Abstract

This invention discloses a temperature control system for aquaculture fermentation beds, belonging to the field of aquaculture fermentation bed temperature control technology. It includes a fermentation bed environment sensing module, a data processing intelligent decision-making module, a precise ventilation control module, an intelligent temperature control module, a humidity collaborative control module, a material turning module, and an anomaly early warning and emergency handling module. This invention constructs an intelligent decision-making system from passive response to active prediction by integrating a multi-parameter fusion algorithm, adaptive PID control, and an LSTM prediction model. Dynamic comprehensive index calculation makes the control strategy closer to the needs of microbial ecology. The improved PID algorithm achieves parameter self-tuning through gradient descent, significantly improving control accuracy and stability. The LSTM model predicts future environmental changes based on historical data, combined with a dynamic threshold configuration unit, enabling the system to automatically optimize control parameters according to the aquaculture stage and raw material ratio, achieving intelligent adaptation across scenarios.

Description

Technical Field

[0001] This invention belongs to the field of temperature regulation technology for aquaculture fermentation beds, specifically referring to a temperature regulation system for aquaculture fermentation beds. Background Technology

[0002] In traditional fermentation bed aquaculture, temperature regulation relies heavily on manual experience or simple temperature control equipment. This means that only temperature and humidity are monitored, without considering key ecological parameters such as oxygen and ammonia concentrations. The control method is a threshold-triggered response, which has a strong lag and is prone to causing environmental fluctuations. Each regulation module operates independently, lacking a coordination mechanism, resulting in high energy consumption and poor performance. Furthermore, the system lacks self-diagnosis and emergency handling capabilities, leading to low system reliability.

[0003] However, current temperature control systems for aquaculture fermentation beds have certain shortcomings. Existing technologies employ a passive control mode based on sensing and response, only taking action after environmental parameters exceed limits, resulting in significant regulatory lag. Furthermore, each control module often operates independently, lacking coordination. Ventilation and temperature control are mostly implemented using "on / off" or segmented control, such as high and low speeds for fans and full on / off operation for heating equipment. This crude control easily causes drastic fluctuations in environmental parameters, which is detrimental to the stability of the microbial community and also wastes energy. Relying on current real-time data makes it impossible to predict future environmental trends. Therefore, preventative measures cannot be taken before temperatures spike or oxygen depletes, leaving the system in a reactive, reactive state. Parameter thresholds are usually fixed and cannot be automatically adjusted according to different breeding stages or different bedding material ratios, requiring frequent manual intervention. The system has low intelligence and is prone to errors. Therefore, this paper proposes a temperature control system for aquaculture fermentation beds. Summary of the Invention

[0004] The purpose of this invention is to provide a temperature control system for aquaculture fermentation beds to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a temperature regulation system for aquaculture fermentation beds, comprising a fermentation bed environment sensing module, a data processing intelligent decision-making module, a precision ventilation regulation module, an intelligent temperature regulation module, a humidity coordinated regulation module, a material turning module, and an abnormal early warning and emergency handling module;

[0006] The fermentation bed environment sensing module is used to collect multi-dimensional environmental parameters of the fermentation bed and transmit them to the data processing intelligent decision-making module.

[0007] The data processing intelligent decision-making module analyzes the received environmental parameters and generates control commands, which are then sent to the precision ventilation adjustment module, intelligent temperature control module, humidity coordination adjustment module, and material turning module, respectively.

[0008] The abnormal early warning and emergency handling module is used to establish real-time communication, obtain the operating status data of each module, and perform early warning and emergency operations.

[0009] Preferably, the fermentation bed environment sensing module includes a distributed temperature sensor array, a humidity sensor, an oxygen concentration sensor, an ammonia concentration sensor, and a data acquisition unit;

[0010] The distributed temperature sensor array uses digital temperature sensors of preset accuracy, arranged in a zigzag pattern on the surface, middle and bottom layers of the fermentation bed, with the spacing between sensors in each layer not exceeding the preset spacing.

[0011] The humidity sensor measures a preset humidity range, which corresponds to the location of the distributed temperature sensor array.

[0012] The oxygen concentration sensor has a preset oxygen concentration measurement range, and the ammonia concentration sensor has a preset ammonia concentration measurement range. They are evenly distributed in the middle layer of the central area of ​​the fermentation bed.

[0013] The data acquisition unit uses a microcontroller with integrated Wi-Fi and 4G communication modules. It has built-in data filtering to preprocess the raw data collected by each sensor by reducing noise and removing outliers, and then transmits the preprocessed data to the data processing intelligent decision-making module.

[0014] Preferably, the data processing intelligent decision-making module includes a local control server, a core algorithm library, and a local database;

[0015] The local control server has a CPU frequency of no less than a preset frequency and a memory capacity of no less than a preset memory capacity, and is used to run the core algorithm library and system management program;

[0016] The core algorithm library includes a multi-parameter fusion algorithm, an improved PID control algorithm, and an LSTM prediction model;

[0017] The multi-parameter fusion algorithm converts the temperature, humidity, oxygen concentration, and ammonia concentration parameters transmitted by the fermentation bed environment sensing module into a comprehensive fermentation bed environment index through a weighted summation method. The weighting coefficients for temperature, humidity, oxygen concentration, and ammonia concentration are preset as follows: ,

[0018] In the formula, F represents the comprehensive environmental index of the fermentation bed, T represents the temperature of the day, H represents the current humidity, O represents the current oxygen concentration, and A represents the current ammonia concentration. Indicates the weight of each parameter;

[0019] The improved PID control algorithm uses the gradient descent method to self-tune the proportional coefficient, integral time constant, and derivative time constant, and the self-tuning period is a preset self-tuning period, as follows: ,

[0020] In the formula, This indicates the k-th control output. This represents the error of the kth iteration. Represents the proportionality coefficient. Represents the integral coefficient. Represents the differential coefficient. Indicates the sampling period. Indicates the coefficient of inertia;

[0021] The LSTM prediction model is trained based on historical environmental parameter data for a preset historical data period and is used to predict the temperature change curve of the fermentation bed within a preset prediction period in the future. The implementation is as follows: ,

[0022] In the formula, This represents the predicted temperature at a future time point t+1. Represents the weight matrix. represents the hidden state of the LSTM at time t, and b represents the bias term;

[0023] The local database uses a MySQL database to store environmental parameter data and the operation logs of each module for a preset storage period.

[0024] Preferably, the precision ventilation control module includes a variable frequency circulating fan, an intelligent air valve, and a wind speed sensor;

[0025] The rated power of the variable frequency circulating fan is a preset rated power, and the speed can be steplessly adjusted within a preset speed range. Each fan corresponds to a preset control area of ​​the fermentation bed, and a guide plate is installed at the fan outlet.

[0026] The intelligent air valve is an electric butterfly valve with an opening adjustment range of a preset opening range, and it corresponds to the variable frequency circulating fan to control the airflow direction at the fan outlet.

[0027] The wind speed sensor is installed at the air outlet of the fan to collect the actual air speed and feed it back to the data processing intelligent decision-making module.

[0028] When the data processing intelligent decision-making module determines whether the temperature of a certain area of ​​the fermentation bed exceeds the preset temperature threshold, it combines the comprehensive index of the fermentation bed environment output by the multi-parameter fusion algorithm. If the temperature exceeds the standard and the ammonia concentration in the comprehensive index is synchronously abnormal, high-speed ventilation is triggered. If the temperature only slightly exceeds the threshold, low-speed ventilation is triggered.

[0029] When the data processing intelligent decision-making module sends a speed adjustment command to the variable frequency circulating fan, it uses the improved PID control algorithm to self-tune the speed parameters based on the difference between the actual temperature and the target temperature, the historical temperature difference accumulation value, and the temperature difference change rate.

[0030] If the LSTM prediction model predicts that the temperature will exceed the standard within a preset prediction period, the data processing intelligent decision-making module sends a low-speed pre-ventilation command to the corresponding area fan in advance to achieve predictive adjustment.

[0031] Preferably, the intelligent temperature control module includes a heating unit, a cooling unit, and a temperature feedback sensor;

[0032] The cooling unit uses an evaporative air cooler with a rated air volume of a preset rated air volume. The air outlet of the air cooler is connected to the air diffuser above the fermentation bed through an air duct. An electric air volume regulating valve is installed in the air duct.

[0033] The temperature feedback sensor is deployed in the working area of ​​the heating unit and the cooling unit to collect temperature data after temperature adjustment in real time and transmit it to the data processing intelligent decision module.

[0034] When determining whether to start temperature adjustment, the data processing intelligent decision-making module combines the multi-parameter fusion algorithm. If the temperature is close to the target after ventilation but the humidity is lower than the preset lower limit, the temperature adjustment is temporarily suspended and the humidity collaborative adjustment module is activated.

[0035] When the data processing intelligent decision-making module sends power adjustment commands to the heating unit and airflow adjustment commands to the cooling unit, it dynamically adjusts the power and airflow based on the temperature feedback sensor data through the improved PID control algorithm.

[0036] If the LSTM prediction model predicts that the temperature will be lower than the preset lower limit within a preset prediction period, the data processing intelligent decision module will preheat the heating unit to standby state in advance; if the predicted temperature will be higher than the upper limit, the cooling unit will be pre-started in advance.

[0037] Preferably, the humidity collaborative adjustment module includes an intelligent spray system, a flow sensor, and a humidity feedback sensor;

[0038] The flow sensor is installed at the water inlet end of the spray pipe to collect spray flow data;

[0039] The humidity feedback sensor and the humidity sensor of the fermentation bed environment sensing module are located in the same position to collect humidity data after spraying.

[0040] When the data processing intelligent decision-making module determines whether to send a spray command, it combines the multi-parameter fusion algorithm to determine whether the humidity is below the lower limit but the temperature is also low. It prioritizes raising the temperature through the heating unit before assessing the spray demand.

[0041] When the data processing intelligent decision-making module adjusts the spray duration and flow rate, it dynamically optimizes the spray parameters by using the improved PID control algorithm with the difference between the target humidity and the actual humidity as input.

[0042] If the LSTM prediction model predicts that the humidity will drop significantly after ventilation, the data processing intelligent decision-making module sends a pre-spray command to the intelligent spray system in advance, and then starts low-flow spraying simultaneously after ventilation is started.

[0043] Preferably, the material turning module includes an automatic turning machine, a limit switch, and a compaction sensor;

[0044] The limit switches are respectively installed at both ends of the fermentation bed to limit the walking boundary of the automatic turning machine. When the turning machine triggers the limit switch, it automatically switches the walking direction.

[0045] The compaction sensor is installed in the middle layer of the fermentation bed to collect compaction data of the bedding material and transmit it to the data processing intelligent decision-making module.

[0046] When the data processing intelligent decision-making module determines whether to send a turning instruction, it combines the multi-parameter fusion algorithm to prioritize increasing the oxygen level through the ventilation module if the oxygen concentration is slightly low but the temperature gradient and compaction are normal, without triggering the turning instruction.

[0047] If the LSTM prediction model predicts that the oxygen concentration will be lower than the preset threshold or the temperature gradient will exceed the limit within a preset prediction period, the data processing intelligent decision-making module will send a turning instruction to the automatic turning machine in advance.

[0048] Preferably, the abnormal early warning and emergency handling module includes an audible and visual alarm, an SMS alarm module, an APP push module, and an emergency control unit;

[0049] The alarm volume of the sound and light alarm is not lower than the preset alarm volume, the flashing frequency of the light is the preset flashing frequency, and it is installed in a conspicuous position in the fermentation bed operation room.

[0050] The SMS alarm module uses a GSM communication module and can send alarm SMS messages to a preset number of administrator mobile phone numbers;

[0051] The APP push module establishes a connection with the administrator's mobile APP via the MQTT protocol to send alarm push information;

[0052] The emergency control unit is a PLC controller independent of the data processing intelligent decision-making module, and it has pre-stored emergency operation programs.

[0053] When the abnormal warning and emergency handling module detects that the sensor data exceeds the normal range and continues for a preset abnormal duration, detects that the execution module does not respond to the control command, or detects that the fermentation bed temperature exceeds the preset upper temperature limit or falls below the preset lower temperature limit, it triggers the corresponding level of warning, and the emergency control unit performs backup sensor switching, backup execution module startup, or emergency shutdown operations according to the pre-stored program.

[0054] Preferably, the distributed temperature sensor array of the fermentation bed environment sensing module is configured with a sensor health diagnosis unit;

[0055] The sensor health diagnostic unit performs self-tests on all temperature sensors during the preset self-test period each day. By sending calibration signals to the sensors and receiving feedback values, it determines whether the sensors are in normal working condition.

[0056] When the difference between the feedback value of a sensor and the feedback value of an adjacent sensor exceeds a preset difference threshold, and the preset number of consecutive samplings exceeds the preset difference threshold, the sensor is determined to be faulty, automatically marked as faulty, and switched to a backup sensor. At the same time, a sensor fault signal is sent to the data processing intelligent decision-making module.

[0057] Preferably, the data processing intelligent decision-making module further includes a dynamic threshold configuration unit;

[0058] The dynamic threshold configuration unit has a built-in parameter threshold table corresponding to different breeding stages and different fermentation raw material ratios;

[0059] Administrators can select the current breeding stage and raw material ratio through the system management program. The dynamic threshold configuration unit will automatically call the corresponding threshold table and update the preset thresholds for temperature, humidity, oxygen concentration and ammonia concentration.

[0060] When the breeding stage changes or the raw material ratio is adjusted, the dynamic threshold configuration unit completes the threshold update within the preset threshold update time and sends the threshold update instruction to all related modules.

[0061] Compared with the prior art, the beneficial effects of the present invention are:

[0062] 1. This invention constructs an intelligent decision-making system that moves from passive response to active prediction by integrating multi-parameter fusion algorithms, adaptive PID control, and LSTM prediction models. Dynamic comprehensive index calculation makes the control strategy more closely aligned with the needs of microbial ecology. The improved PID algorithm achieves parameter self-tuning through gradient descent, significantly improving control accuracy and stability. The LSTM model predicts future environmental changes based on historical data, and combined with a dynamic threshold configuration unit, enables the system to automatically optimize control parameters according to the breeding stage and raw material ratio, achieving intelligent adaptation across scenarios.

[0063] 2. This invention achieves refined management of ventilation volume through a zoned air control and intelligent air valve linkage design, and through variable frequency fan speed adjustment and electric butterfly valve angle control; combined with the predictive control capability of the LSTM predictive model, preventive ventilation can be initiated before environmental parameters become abnormal, effectively avoiding sudden temperature rises or local over-blowing problems; the closed-loop feedback mechanism ensures uniform airflow distribution, reducing energy waste while improving control efficiency.

[0064] 3. This invention achieves precise and energy-saving temperature regulation through the complementary design of heating and cooling units, combined with real-time monitoring by temperature feedback sensors; the dynamic power regulation strategy avoids temperature oscillations caused by traditional full-on and full-off modes, and the pre-start mechanism driven by the LSTM prediction model significantly shortens the temperature recovery time; the linkage control with the humidity co-regulation module effectively avoids the risks of dry heat or humid cold environments.

[0065] 4. This invention solves the problems of lag and over-adjustment in traditional humidity regulation by using a coordinated strategy of spray system and ventilation and temperature control modules and closed-loop control of flow and humidity feedback; the predictive pre-spray mechanism combined with LSTM model can respond in advance to humidity fluctuations caused by ventilation and keep the moisture content of the bedding material within the optimal range for microbial activity; the modular design significantly reduces the proportion of spray energy consumption, while parametric control avoids bedding material clumping.

[0066] 5. This invention achieves intelligent decision-making for mechanical turning through a compaction monitoring and predictive turning mechanism; the tracked turning machine dynamically adjusts the turning depth and frequency according to the temperature gradient and microbial oxygen demand, avoiding the damage to the microbial community caused by traditional timed turning; the predictive ability of the LSTM model can trigger the turning operation in advance, maintaining the looseness of the bedding material and the balance of oxygen supply, while optimizing the walking speed through humidity parameters, effectively improving the turning efficiency and bedding material homogeneity. Attached Figure Description

[0067] Figure 1 This is a schematic diagram of the structure of a temperature regulation system for a livestock fermentation bed according to the present invention;

[0068] Figure 2 The present invention provides the operating flow of a temperature control system for a livestock fermentation bed. Figure 1 ;

[0069] Figure 3 The present invention provides the operating flow of a temperature control system for a livestock fermentation bed. Figure 2 . Detailed Implementation

[0070] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Example

[0071] Please see Figures 1-3 As shown, the present invention provides a technical solution including a fermentation bed environment sensing module, a data processing intelligent decision-making module, a precise ventilation adjustment module, an intelligent temperature control module, a humidity coordinated adjustment module, a material turning module, and an abnormal early warning and emergency handling module;

[0072] The fermentation bed environment sensing module is used to collect multi-dimensional environmental parameters of the fermentation bed and transmit them to the data processing intelligent decision-making module.

[0073] The data processing intelligent decision-making module analyzes the received environmental parameters and generates control commands, which are then sent to the precision ventilation adjustment module, intelligent temperature control module, humidity coordination adjustment module, and material turning module, respectively.

[0074] The abnormal early warning and emergency handling module is used to establish real-time communication, obtain the operating status data of each module, and perform early warning and emergency operations.

[0075] In this embodiment, the fermentation bed environment sensing module includes a distributed temperature sensor array, a humidity sensor, an oxygen concentration sensor, an ammonia concentration sensor, and a data acquisition device;

[0076] The distributed temperature sensor array uses digital temperature sensors of preset precision, arranged in a zigzag path on the surface of the fermentation bed, with preset surface depth, middle layer, middle layer depth and bottom layer depth, and the sensor spacing in each layer is no greater than the preset spacing.

[0077] The humidity sensor measures a preset humidity range, which corresponds to the location of the distributed temperature sensor array.

[0078] The oxygen concentration sensor has a preset oxygen concentration measurement range, and the ammonia concentration sensor has a preset ammonia concentration measurement range. They are evenly distributed in the middle layer of the central area of ​​the fermentation bed.

[0079] The data acquisition unit uses a microcontroller with integrated Wi-Fi and 4G communication modules. It has built-in data filtering to preprocess the raw data collected by each sensor by reducing noise and removing outliers, and then transmits the preprocessed data to the data processing intelligent decision-making module.

[0080] In this embodiment, the data processing intelligent decision-making module includes a local control server, a core algorithm library, and a local database;

[0081] The local control server has a CPU frequency of no less than a preset frequency and a memory capacity of no less than a preset memory capacity, and is used to run the core algorithm library and system management program;

[0082] The core algorithm library includes a multi-parameter fusion algorithm, an improved PID control algorithm, and an LSTM prediction model;

[0083] The multi-parameter fusion algorithm converts the temperature, humidity, oxygen concentration, and ammonia concentration parameters transmitted by the fermentation bed environment sensing module into a comprehensive fermentation bed environment index through a weighted summation method. The weighting coefficients for temperature, humidity, oxygen concentration, and ammonia concentration are preset as follows: ,

[0084] In the formula, F represents the comprehensive environmental index of the fermentation bed, T represents the temperature of the day, H represents the current humidity, O represents the current oxygen concentration, and A represents the current ammonia concentration. Indicates the weight of each parameter;

[0085] The improved PID control algorithm uses the gradient descent method to self-tune the proportional coefficient, integral time constant, and derivative time constant, and the self-tuning period is a preset self-tuning period, as follows: ,

[0086] In the formula, This indicates the k-th control output. This represents the error of the kth iteration. Represents the proportionality coefficient. Represents the integral coefficient. Represents the differential coefficient. Indicates the sampling period. Indicates the coefficient of inertia;

[0087] The LSTM prediction model is trained based on historical environmental parameter data for a preset historical data period and is used to predict the temperature change curve of the fermentation bed within a preset prediction period in the future. The implementation is as follows: ,

[0088] In the formula, This represents the predicted temperature at a future time point t+1. This represents the weight matrix, which contains the weights of the output layer. This represents the hidden state of the LSTM at time t. for , for b represents the bias term;

[0089] The local database uses a MySQL database to store environmental parameter data and the operation logs of each module for a preset storage period.

[0090] In this embodiment, the precision ventilation control module includes a variable frequency circulating fan, an intelligent air valve, and a wind speed sensor;

[0091] The rated power of the variable frequency circulating fan is a preset rated power, and the speed can be steplessly adjusted within a preset speed range. Each fan corresponds to a preset control area of ​​the fermentation bed, and a guide plate is installed at the fan outlet.

[0092] The intelligent air valve is an electric butterfly valve with an opening adjustment range of a preset opening range, and it corresponds to the variable frequency circulating fan to control the airflow direction at the fan outlet.

[0093] The wind speed sensor is installed at the air outlet of the fan to collect the actual air speed and feed it back to the data processing intelligent decision-making module.

[0094] When the data processing intelligent decision-making module determines whether the temperature of a certain area of ​​the fermentation bed exceeds the preset temperature threshold, it combines the comprehensive index of the fermentation bed environment output by the multi-parameter fusion algorithm. If the temperature exceeds the standard and the ammonia concentration in the comprehensive index is synchronously abnormal, high-speed ventilation is triggered. If the temperature only slightly exceeds the threshold, low-speed ventilation is triggered.

[0095] When the data processing intelligent decision-making module sends a speed adjustment command to the variable frequency circulating fan, it uses the improved PID control algorithm to self-tune the speed parameters based on the difference between the actual temperature and the target temperature, the historical temperature difference accumulation value, and the temperature difference change rate, so as to avoid sudden speed changes.

[0096] If the LSTM prediction model predicts that the temperature will exceed the standard within a preset prediction period, the data processing intelligent decision-making module sends a low-speed pre-ventilation command to the corresponding area fan in advance to achieve predictive adjustment.

[0097] In this embodiment, the intelligent temperature control module includes a heating unit, a cooling unit, and a temperature feedback sensor;

[0098] The heating unit uses carbon fiber far-infrared heating plates. The power of a single heating plate is a preset heating power, and they are arranged at the bottom of the fermentation bed at a density of one plate per preset area. The heating plate is wrapped with a high-temperature resistant insulating layer.

[0099] The cooling unit uses an evaporative air cooler with a rated air volume of a preset rated air volume. The air outlet of the air cooler is connected to the air diffuser above the fermentation bed through an air duct. An electric air volume regulating valve is installed in the air duct.

[0100] The temperature feedback sensor is deployed in the working area of ​​the heating unit and the cooling unit to collect temperature data after temperature adjustment in real time and transmit it to the data processing intelligent decision module.

[0101] When determining whether to start temperature adjustment, the data processing intelligent decision-making module combines the multi-parameter fusion algorithm. If the temperature is close to the target after ventilation but the humidity is lower than the preset lower limit, the temperature adjustment is temporarily suspended and the humidity collaborative adjustment module is activated.

[0102] When the data processing intelligent decision-making module sends power adjustment commands to the heating unit and airflow adjustment commands to the cooling unit, it dynamically adjusts the power and airflow based on the temperature feedback sensor data through the improved PID control algorithm to avoid sudden temperature rises and falls.

[0103] If the LSTM prediction model predicts that the temperature will be lower than the preset lower limit within a preset prediction period, the data processing intelligent decision module will preheat the heating unit to standby state in advance; if the predicted temperature will be higher than the upper limit, the cooling unit will be pre-started in advance.

[0104] In this embodiment, the humidity collaborative adjustment module includes an intelligent spray system, a flow sensor, and a humidity feedback sensor;

[0105] The intelligent spray system consists of a high-pressure water pump, spray pipes, and atomizing nozzles. The rated pressure of the high-pressure water pump is a preset water pump pressure. The spray pipes are laid along the length of the fermentation bed. The spacing of the atomizing nozzles is a preset nozzle spacing, and the spray angle of the nozzles is adjustable.

[0106] The flow sensor is installed at the water inlet end of the spray pipe to collect spray flow data;

[0107] The humidity feedback sensor and the humidity sensor of the fermentation bed environment sensing module are located in the same position to collect humidity data after spraying.

[0108] When the data processing intelligent decision-making module determines whether to send a spray command, it combines the multi-parameter fusion algorithm to determine whether the humidity is below the lower limit but the temperature is also low. It prioritizes raising the temperature through the heating unit before assessing the spray demand.

[0109] When the data processing intelligent decision-making module adjusts the spray duration and flow rate, it uses the improved PID control algorithm to dynamically optimize the spray parameters with the difference between the target humidity and the actual humidity as input, so as to avoid over-adjustment of humidity.

[0110] If the LSTM prediction model predicts that the humidity will drop significantly after ventilation, the data processing intelligent decision-making module sends a pre-spray command to the intelligent spray system in advance, and then starts low-flow spraying simultaneously after ventilation is started.

[0111] In this embodiment, the material turning module includes an automatic turning and turning machine, a limit switch, and a compaction sensor;

[0112] The automatic turning and turning machine adopts a tracked walking mechanism. The diameter of the turning and turning toothed roller is a preset toothed roller diameter, the turning and turning depth can be adjusted within a preset turning and turning depth range, and the walking speed is a preset walking speed.

[0113] The limit switches are respectively installed at both ends of the fermentation bed to limit the walking boundary of the automatic turning machine. When the turning machine triggers the limit switch, it automatically switches the walking direction.

[0114] The compaction sensor is installed in the middle layer of the fermentation bed to collect compaction data of the bedding material and transmit it to the data processing intelligent decision-making module.

[0115] When the data processing intelligent decision-making module determines whether to send a turning instruction, it combines the multi-parameter fusion algorithm to prioritize increasing the oxygen level through the ventilation module if the oxygen concentration is slightly low but the temperature gradient and compaction are normal, without triggering the turning instruction.

[0116] If the LSTM prediction model predicts that the oxygen concentration will be lower than the preset threshold or the temperature gradient will exceed the standard within the preset prediction period, the data processing intelligent decision-making module will send a turning instruction to the automatic turning machine in advance to avoid parameter deterioration.

[0117] When the data processing intelligent decision-making module adjusts the turning depth and walking speed, it refers to the weight of the humidity parameter in the multi-parameter fusion algorithm. When the humidity is too high, the walking speed is reduced to avoid the bedding material from clumping.

[0118] In this embodiment, the abnormal early warning and emergency handling module includes an audible and visual alarm, an SMS alarm module, an APP push module, and an emergency control unit;

[0119] The alarm volume of the sound and light alarm is not lower than the preset alarm volume, the flashing frequency of the light is the preset flashing frequency, and it is installed in a conspicuous position in the fermentation bed operation room.

[0120] The SMS alarm module uses a GSM communication module and can send alarm SMS messages to a preset number of administrator mobile phone numbers;

[0121] The APP push module establishes a connection with the administrator's mobile APP via the MQTT protocol to send alarm push information;

[0122] The emergency control unit is a PLC controller independent of the data processing intelligent decision-making module, and it has pre-stored emergency operation programs.

[0123] When the abnormal warning and emergency handling module detects that sensor data exceeds the normal range and continues for a preset abnormal duration, detects that a certain execution module does not respond to control commands, or detects that the fermentation bed temperature exceeds the preset upper temperature limit or falls below the preset lower temperature limit, it triggers the corresponding level of warning, and the emergency control unit performs backup sensor switching, backup execution module startup, or emergency shutdown operations according to the pre-stored program.

[0124] In this embodiment, the distributed temperature sensor array of the fermentation bed environment sensing module is configured with a sensor health diagnosis unit;

[0125] The sensor health diagnostic unit performs self-tests on all temperature sensors during the preset self-test period each day. By sending calibration signals to the sensors and receiving feedback values, it determines whether the sensors are in normal working condition.

[0126] When the difference between the feedback value of a sensor and the feedback value of an adjacent sensor exceeds a preset difference threshold, and the preset number of consecutive samplings exceeds the preset difference threshold, the sensor is determined to be faulty, automatically marked as faulty, and switched to a backup sensor. The number of backup sensors is a preset ratio of the number of main sensors, and a sensor fault signal is sent to the data processing intelligent decision-making module.

[0127] Preferably, the data processing intelligent decision-making module further includes a dynamic threshold configuration unit;

[0128] The dynamic threshold configuration unit has a built-in parameter threshold table corresponding to different breeding stages and different fermentation raw material ratios;

[0129] Administrators can select the current breeding stage and raw material ratio through the system management program. The dynamic threshold configuration unit will automatically call the corresponding threshold table and update the preset thresholds for temperature, humidity, oxygen concentration and ammonia concentration.

[0130] When the breeding stage changes or the raw material ratio is adjusted, the dynamic threshold configuration unit completes the threshold update within the preset threshold update time and sends the threshold update instruction to all related modules.

[0131] Working Principle: The system comprehensively collects environmental data from the fermentation bed. It utilizes a three-dimensional monitoring network consisting of temperature sensor arrays distributed across the surface, middle, and bottom layers of the fermentation bed, along with corresponding humidity, oxygen, and ammonia concentration sensors. All sensor data is collected by a data acquisition unit with integrated wireless communication capabilities. The acquisition unit first preprocesses the raw data, performing noise reduction and outlier removal, before transmitting reliable data to the decision-making module. The intelligent decision-making module receives the environmental data and uses three core algorithms for intelligent decision-making: First, a multi-parameter fusion algorithm weights and calculates multiple environmental parameters into a comprehensive index for overall evaluation; second, an improved PID control algorithm enables the system to adaptively tune the control. The system optimizes parameters to achieve precise and smooth control output, avoiding overshoot in actuator movements. Finally, it uses an LSTM prediction model to predict future temperature trends based on historical data, thus transitioning from passive response to proactive prediction. All data and logs are stored in a local database. Ventilation of the fermentation bed is adjusted according to instructions from the decision module. When the temperature in a certain area exceeds the standard, the system combines parameters such as ammonia concentration to determine whether to use high or low speed ventilation. During ventilation, the system dynamically adjusts the speed of the variable frequency fan and the opening of the intelligent air valve using a PID algorithm to ensure smooth ventilation. Furthermore, when the system predicts that the temperature will exceed the standard in the future, it will instruct the fan to operate at low speed in advance. Ventilation; precise temperature control through both heating and cooling; heating units utilize carbon fiber heating plates installed at the bottom of the bed, while cooling units employ evaporative air coolers; the system performs a comprehensive assessment before initiating temperature adjustment, for example, if humidity is already low, it will temporarily postpone any cooling measures that might further reduce humidity; during temperature adjustment, the heating power or cooling airflow is dynamically adjusted based on a PID algorithm to ensure a smooth temperature transition; the system also preheats or pre-starts the heating or cooling units to standby mode based on predictions; a temperature coordination module maintains suitable humidity in the fermentation bed through a high-pressure spray system; the system's decision-making process fully considers the linkage with other parameters, such as when humidity and temperature are both... When the humidity is low, heating will be activated first, as the increase in temperature helps to raise humidity, thus avoiding indiscriminate spraying that could lead to environmental degradation. The duration and flow rate of the spraying are precisely controlled by a PID algorithm. The system can also predict the humidity drop that ventilation may cause, and prepare in advance by activating low-flow spraying simultaneously when ventilation begins to compensate. An automatic turning machine is used to improve the physical structure and air permeability of the bedding material. The system does not turn the bedding material just because the oxygen concentration is slightly low, but rather comprehensively evaluates the temperature gradient and the usual compaction data, prioritizing ventilation as a solution, with turning the bedding material as a more thorough backup. The system uses a predictive model to trigger the turning command in advance before it predicts that there will be insufficient oxygen or an imbalance in the temperature gradient.The turning depth and speed are adjusted according to real-time humidity. When humidity is high, the speed is reduced to prevent bedding material from clumping. An abnormal early warning and emergency handling module monitors all sensor data and the status of the execution modules. If data exceeds limits, equipment malfunctions, or temperatures become drastically abnormal, it will issue alarms through multiple channels, including sound and light, SMS, and a mobile app, based on preset levels. Simultaneously, its built-in independent emergency control unit will immediately take over the system, executing pre-stored plans such as switching to backup sensors, starting backup equipment, or emergency shutdown. It periodically performs self-checks on all temperature sensors by sending calibration signals and comparing their feedback values ​​with those of adjacent sensors to determine if the sensors are functioning correctly. If a sensor is determined to be faulty, the system will automatically isolate it and switch to a backup sensor, while simultaneously reporting the fault information. It can adapt to different breeding stages and feed ratios. It internally stores parameter threshold tables corresponding to different situations. When the administrator selects or switches breeding stages and feeds, this unit will automatically update the control thresholds for temperature, humidity, and gas concentration used by the system and notify all relevant modules, ensuring that the overall system control objectives remain consistent with current actual production needs.

[0132] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their likenesses.

[0133] The present invention and its embodiments have been described above. This description is not restrictive, and the accompanying drawings are only one embodiment of the present invention; the actual structure is not limited thereto. In conclusion, if those skilled in the art are inspired by this description and design similar structures and embodiments without departing from the spirit of the invention, such designs should fall within the protection scope of the present invention.

Claims

1. A temperature control system for aquaculture fermentation beds, characterized in that: It includes a fermentation bed environment sensing module, a data processing intelligent decision-making module, a precise ventilation adjustment module, an intelligent temperature control module, a humidity coordinated adjustment module, a material turning module, and an abnormal early warning and emergency handling module; The fermentation bed environment sensing module is used to collect multi-dimensional environmental parameters of the fermentation bed and transmit them to the data processing intelligent decision-making module. The data processing intelligent decision-making module analyzes the received environmental parameters and generates control commands, which are then sent to the precision ventilation adjustment module, intelligent temperature control module, humidity coordination adjustment module, and material turning module, respectively. The abnormal early warning and emergency handling module is used to establish real-time communication, obtain the operating status data of each module, and perform early warning and emergency operations.

2. The temperature control system for aquaculture fermentation beds according to claim 1, characterized in that: The fermentation bed environment sensing module includes a distributed temperature sensor array, a humidity sensor, an oxygen concentration sensor, an ammonia concentration sensor, and a data acquisition unit. The distributed temperature sensor array uses digital temperature sensors of preset accuracy, arranged in a zigzag pattern on the surface, middle and bottom layers of the fermentation bed, with the spacing between sensors in each layer not exceeding the preset spacing. The humidity sensor measures a preset humidity range, which corresponds to the location of the distributed temperature sensor array. The oxygen concentration sensor has a preset oxygen concentration measurement range, and the ammonia concentration sensor has a preset ammonia concentration measurement range. They are evenly distributed in the middle layer of the central area of ​​the fermentation bed. The data acquisition unit uses a microcontroller with integrated Wi-Fi and 4G communication modules. It has built-in data filtering to preprocess the raw data collected by each sensor by reducing noise and removing outliers, and then transmits the preprocessed data to the data processing intelligent decision-making module.

3. The temperature control system for aquaculture fermentation beds according to claim 1, characterized in that: The data processing intelligent decision-making module includes a local control server, a core algorithm library, and a local database. The local control server has a CPU frequency of no less than a preset frequency and a memory capacity of no less than a preset memory capacity, and is used to run the core algorithm library and system management program; The core algorithm library includes a multi-parameter fusion algorithm, an improved PID control algorithm, and an LSTM prediction model; The multi-parameter fusion algorithm converts the temperature, humidity, oxygen concentration, and ammonia concentration parameters transmitted by the fermentation bed environment sensing module into a comprehensive fermentation bed environment index through a weighted summation method. The weighting coefficients for temperature, humidity, oxygen concentration, and ammonia concentration are preset as follows: , In the formula, F represents the comprehensive environmental index of the fermentation bed, T represents the temperature of the day, H represents the current humidity, O represents the current oxygen concentration, and A represents the current ammonia concentration. Indicates the weight of each parameter; The improved PID control algorithm uses the gradient descent method to self-tune the proportional coefficient, integral time constant, and derivative time constant, and the self-tuning period is a preset self-tuning period, as follows: , In the formula, This indicates the k-th control output. This represents the error of the kth iteration. Represents the proportionality coefficient. Represents the integral coefficient. Represents the differential coefficient. Indicates the sampling period. Indicates the coefficient of inertia; The LSTM prediction model is trained based on historical environmental parameter data for a preset historical data period and is used to predict the temperature change curve of the fermentation bed within a preset prediction period in the future. The implementation is as follows: , In the formula, This represents the predicted temperature at a future time point t+1. Represents the weight matrix. represents the hidden state of the LSTM at time t, and b represents the bias term; The local database uses a MySQL database to store environmental parameter data and the operation logs of each module for a preset storage period.

4. The temperature control system for aquaculture fermentation beds according to claim 1, characterized in that: The precision ventilation control module includes a variable frequency circulating fan, an intelligent air valve, and a wind speed sensor. The rated power of the variable frequency circulating fan is a preset rated power, and the speed can be steplessly adjusted within a preset speed range. Each fan corresponds to a preset control area of ​​the fermentation bed, and a guide plate is installed at the fan outlet. The intelligent air valve is an electric butterfly valve with an opening adjustment range of a preset opening range, and it corresponds to the variable frequency circulating fan to control the airflow direction at the fan outlet. The wind speed sensor is installed at the air outlet of the fan to collect the actual air speed and feed it back to the data processing intelligent decision-making module. When the data processing intelligent decision-making module determines whether the temperature of a certain area of ​​the fermentation bed exceeds the preset temperature threshold, it combines the comprehensive index of the fermentation bed environment output by the multi-parameter fusion algorithm. If the temperature exceeds the standard and the ammonia concentration in the comprehensive index is synchronously abnormal, high-speed ventilation is triggered. If the temperature only slightly exceeds the threshold, low-speed ventilation is triggered. When the data processing intelligent decision-making module sends a speed adjustment command to the variable frequency circulating fan, it uses the improved PID control algorithm to self-tune the speed parameters based on the difference between the actual temperature and the target temperature, the historical temperature difference accumulation value, and the temperature difference change rate. If the LSTM prediction model predicts that the temperature will exceed the standard within a preset prediction period, the data processing intelligent decision-making module sends a low-speed pre-ventilation command to the corresponding area fan in advance to achieve predictive adjustment.

5. The temperature control system for aquaculture fermentation beds according to claim 1, characterized in that: The intelligent temperature control module includes a heating unit, a cooling unit, and a temperature feedback sensor; The cooling unit uses an evaporative air cooler with a rated air volume of a preset rated air volume. The air outlet of the air cooler is connected to the air diffuser above the fermentation bed through an air duct. An electric air volume regulating valve is installed in the air duct. The temperature feedback sensor is deployed in the working area of ​​the heating unit and the cooling unit to collect temperature data after temperature adjustment in real time and transmit it to the data processing intelligent decision module. When determining whether to start temperature adjustment, the data processing intelligent decision-making module combines the multi-parameter fusion algorithm. If the temperature is close to the target after ventilation but the humidity is lower than the preset lower limit, the temperature adjustment is temporarily suspended and the humidity collaborative adjustment module is activated. When the data processing intelligent decision-making module sends power adjustment commands to the heating unit and airflow adjustment commands to the cooling unit, it dynamically adjusts the power and airflow based on the temperature feedback sensor data through the improved PID control algorithm. If the LSTM prediction model predicts that the temperature will be lower than the preset lower limit within a preset prediction period, the data processing intelligent decision module will preheat the heating unit to standby state in advance; if the predicted temperature will be higher than the upper limit, the cooling unit will be pre-started in advance.

6. The temperature control system for aquaculture fermentation beds according to claim 1, characterized in that: The humidity collaborative adjustment module includes an intelligent spray system, a flow sensor, and a humidity feedback sensor; The flow sensor is installed at the water inlet end of the spray pipe to collect spray flow data; The humidity feedback sensor and the humidity sensor of the fermentation bed environment sensing module are located in the same position to collect humidity data after spraying. When the data processing intelligent decision-making module determines whether to send a spray command, it combines the multi-parameter fusion algorithm to determine whether the humidity is below the lower limit but the temperature is also low. It prioritizes raising the temperature through the heating unit before assessing the spray demand. When the data processing intelligent decision-making module adjusts the spray duration and flow rate, it dynamically optimizes the spray parameters by using the improved PID control algorithm with the difference between the target humidity and the actual humidity as input. If the LSTM prediction model predicts that the humidity will drop significantly after ventilation, the data processing intelligent decision-making module sends a pre-spray command to the intelligent spray system in advance, and then starts low-flow spraying simultaneously after ventilation is started.

7. The temperature control system for aquaculture fermentation beds according to claim 1, characterized in that: The material turning module includes an automatic turning and turning machine, a limit switch, and a compaction sensor; The limit switches are respectively installed at both ends of the fermentation bed to limit the walking boundary of the automatic turning machine. When the turning machine triggers the limit switch, it automatically switches the walking direction. The compaction sensor is installed in the middle layer of the fermentation bed to collect compaction data of the bedding material and transmit it to the data processing intelligent decision-making module. When the data processing intelligent decision-making module determines whether to send a turning instruction, it combines the multi-parameter fusion algorithm to prioritize increasing the oxygen level through the ventilation module if the oxygen concentration is slightly low but the temperature gradient and compaction are normal, without triggering the turning instruction. If the LSTM prediction model predicts that the oxygen concentration will be lower than the preset threshold or the temperature gradient will exceed the limit within a preset prediction period, the data processing intelligent decision-making module will send a turning instruction to the automatic turning machine in advance.

8. The temperature control system for aquaculture fermentation beds according to claim 1, characterized in that: The abnormal early warning and emergency handling module includes an audible and visual alarm, an SMS alarm module, an APP push module, and an emergency control unit; The alarm volume of the sound and light alarm is not lower than the preset alarm volume, the flashing frequency of the light is the preset flashing frequency, and it is installed in a conspicuous position in the fermentation bed operation room. The SMS alarm module uses a GSM communication module and can send alarm SMS messages to a preset number of administrator mobile phone numbers; The APP push module establishes a connection with the administrator's mobile APP via the MQTT protocol to send alarm push information; The emergency control unit is a PLC controller independent of the data processing intelligent decision-making module, and it has pre-stored emergency operation programs. When the abnormal warning and emergency handling module detects that the sensor data exceeds the normal range and continues for a preset abnormal duration, detects that the execution module does not respond to the control command, or detects that the fermentation bed temperature exceeds the preset upper temperature limit or falls below the preset lower temperature limit, it triggers the corresponding level of warning, and the emergency control unit performs backup sensor switching, backup execution module startup, or emergency shutdown operations according to the pre-stored program.

9. The temperature control system for aquaculture fermentation beds according to claim 2, characterized in that: The fermentation bed environment sensing module includes a distributed temperature sensor array configured with a sensor health diagnosis unit. The sensor health diagnostic unit performs self-tests on all temperature sensors during the preset self-test period each day. By sending calibration signals to the sensors and receiving feedback values, it determines whether the sensors are in normal working condition. When the difference between the feedback value of a sensor and the feedback value of an adjacent sensor exceeds a preset difference threshold, and the preset number of consecutive samplings exceeds the preset difference threshold, the sensor is determined to be faulty, automatically marked as faulty, and switched to a backup sensor. At the same time, a sensor fault signal is sent to the data processing intelligent decision-making module.

10. The temperature control system for aquaculture fermentation bed according to claim 3, characterized in that: The data processing intelligent decision-making module also includes a dynamic threshold configuration unit; The dynamic threshold configuration unit has a built-in parameter threshold table corresponding to different breeding stages and different fermentation raw material ratios; Administrators can select the current breeding stage and raw material ratio through the system management program. The dynamic threshold configuration unit will automatically call the corresponding threshold table and update the preset thresholds for temperature, humidity, oxygen concentration and ammonia concentration. When the breeding stage changes or the raw material ratio is adjusted, the dynamic threshold configuration unit completes the threshold update within the preset threshold update time and sends the threshold update instruction to all related modules.

Images