A poultry house environment self-adaptive regulation method and system

Abstract

The application relates to the technical field of poultry breeding, and discloses a poultry house environment self-adaptive regulation method and system. The method comprises the following steps: collecting poultry physiological data through a sensor array to obtain a body weight growth rate curve and a metabolic intensity fluctuation curve, and combining a preset feather development model to determine a physiological demand vector; evaluating the feather development degree of the poultry based on the physiological demand vector to obtain a quantitative score, analyzing the correlation between the quantitative score and the physiological demand vector, and calculating a wind speed adjustment parameter based on the correlation; generating an air flow distribution mode sequence based on the wind speed adjustment parameter, judging whether there is a fluctuation anomaly, obtaining a wind speed control signal, and transmitting the wind speed control signal to a ventilation equipment to adjust the wind speed intensity and the ventilation frequency; collecting air quality feedback data in the poultry house in real time, inputting the air quality feedback data into the feather development model, determining an adjustment value of the physiological demand vector in the next period, and generating a breeding benefit evaluation report according to the adjustment value and dynamic fluctuation data. The application improves the poultry breeding efficiency.

Description

Technical Field

[0001] This application relates to the field of poultry farming technology, and in particular to a method and system for adaptive control of poultry house environment. Background Technology

[0002] In poultry farming, environmental control plays a crucial role in poultry growth and development. Existing poultry house environmental control technologies primarily rely on fixed parameter settings to regulate the poultry rearing environment, a method with several drawbacks. First, traditional technologies typically rely on timed and simple wind speed adjustments to ensure air circulation and temperature and humidity control. However, these methods fail to adapt to the changing needs of poultry at different growth stages, resulting in delayed environmental control. Poultry cannot obtain an environment that adapts in real time, triggering stress responses and impacting their health and productivity. Second, existing technologies often lack real-time dynamic responses to environmental control, failing to fully utilize real-time changes in poultry physiological data (such as weight gain rate and metabolic intensity) and environmental data (such as temperature, humidity, and carbon dioxide concentration). This prevents environmental control from accurately matching the specific needs of poultry and cannot avoid unnecessary resource waste, leading to energy and resource depletion. Furthermore, traditional technologies neglect the complex relationship between environmental factors and poultry physiological needs, failing to provide personalized and precise adjustments to control parameters such as wind speed, temperature, and humidity. This results in over- or under-regulation of the ventilation system, further affecting poultry growth. Furthermore, these systems typically lack adaptability, failing to optimize in real-time based on environmental and poultry physiological changes, and lacking intelligent control capabilities. Simultaneously, existing technologies suffer from insufficient data integration and predictive capabilities. Traditional systems often rely on a single data source, failing to effectively integrate and analyze multidimensional data, resulting in low control precision and poor effects, ultimately impacting farming efficiency and resource utilization. Due to these shortcomings, existing technologies cannot effectively improve poultry house environmental management, leading to significant resource waste and negatively impacting poultry health and growth.

[0003] Therefore, there is an urgent need for an intelligent and adaptive environmental control system to provide more precise and flexible environmental control solutions, thereby improving poultry production efficiency and farming benefits. Summary of the Invention

[0004] To address the aforementioned technical problems, this application provides an adaptive control method and system for poultry house environment, which can improve poultry growth efficiency, reduce resource waste, and optimize breeding benefits.

[0005] In a first aspect, this application provides a method for adaptive control of poultry house environment, the method comprising:

[0006] Step S1: Collect poultry physiological data through a sensor array to obtain a weight gain rate curve and a metabolic intensity fluctuation curve. Based on the weight gain rate curve and the metabolic intensity fluctuation curve, determine the physiological demand vector in conjunction with a preset feather development model.

[0007] Step S2: Evaluate the degree of bird feather development based on the physiological demand vector to obtain a quantitative score, analyze the correlation between the quantitative score and the physiological demand vector, and calculate the wind speed adjustment parameters based on the correlation;

[0008] Step S3: Based on the wind speed adjustment parameters, generate an airflow distribution pattern sequence, determine whether there are any abnormal fluctuations, optimize the wind speed control signal, and transmit the wind speed control signal to the ventilation equipment, which adjusts the wind speed intensity and ventilation frequency.

[0009] Step S4: Collect air quality feedback data in the poultry house in real time, input the air quality feedback data into the feather development model, determine the physiological demand vector adjustment value for the next cycle, and generate a breeding benefit assessment report based on the adjustment value and dynamic fluctuation data.

[0010] In conjunction with the first aspect, in the first implementation of the first aspect of this application, physiological data of poultry are collected through a sensor array to obtain a weight gain rate curve and a metabolic intensity fluctuation curve, including:

[0011] The poultry physiological data collected by the sensor array includes poultry weight data and metabolic index signals. The poultry weight data and metabolic index signals are processed by signal filtering to remove noise interference. Based on the filtered data, the weight gain rate is calculated and a weight gain rate curve is generated. Based on the filtered metabolic index signals, the metabolic intensity change is analyzed and a metabolic intensity fluctuation curve is generated.

[0012] In conjunction with the first aspect, in the second implementation of the first aspect of this application, obtaining the physiological demand vector for the current growth stage includes:

[0013] The feature data of the weight gain rate curve and the metabolic intensity fluctuation curve are extracted by the parameter fusion algorithm. A feather development model is established and trained based on historical poultry growth data and physiological index data. The feature data is input into the feather development model. The feather development model divides the growth process of poultry into multiple growth stages. The current growth stage of poultry is determined based on the threshold of the feature data. The physiological demand vector corresponding to the current growth stage is obtained, including nutritional demand, temperature demand and ventilation demand components.

[0014] In conjunction with the first aspect, in the third implementation of the first aspect of this application, the degree of bird feather development is assessed based on the physiological demand vector to obtain a quantitative score, including:

[0015] Determine whether the physiological demand vector exceeds a preset threshold. If it does, activate the dynamic evaluation module.

[0016] The dynamic evaluation module loads an image acquisition device to acquire images of poultry;

[0017] The feather texture features of the bird images are extracted based on the convolutional neural network algorithm, and the feather coverage density is calculated.

[0018] A quantitative score for the degree of feather development is calculated based on the stated feather coverage density.

[0019] In conjunction with the first aspect, in the fourth implementation of the first aspect of this application, calculating the feather cover density includes:

[0020] The feather region of a bird image is segmented at the pixel level using a semantic segmentation algorithm. The feather coverage area ratio is calculated based on the segmentation results to obtain the feather coverage density.

[0021] In conjunction with the first aspect, in the fifth implementation of the first aspect of this application, the wind speed adjustment parameters are calculated based on correlation, including:

[0022] The quantified score and the physiological requirement vector are mapped to a preset fuzzy set using a fuzzy logic algorithm. Fuzzy reasoning is then performed using a rule base to obtain the correlation value between poultry physiological requirements and environmental regulation parameters.

[0023] The correlation value is multiplied by a preset weighting factor to obtain a preliminary coefficient, and the preliminary coefficient is limited to a preset range by linear interpolation to generate a wind speed intensity adjustment coefficient.

[0024] The air volume transition parameters are generated based on the wind speed intensity adjustment coefficient, wherein the air volume transition parameters gradually transition from the current air volume to the target air volume in a sequence.

[0025] The airflow transition parameters are smoothed to obtain stable airflow adjustment parameters.

[0026] In conjunction with the first aspect, in the sixth implementation of the first aspect of this application, based on the wind speed adjustment parameters, an airflow distribution pattern sequence is generated, and it is determined whether there are any fluctuation anomalies, including:

[0027] The wind speed adjustment parameters are input into the sequence generation function to calculate the initial airflow distribution pattern sequence;

[0028] Calculate the wind speed difference between adjacent time points in the initial airflow distribution pattern sequence to obtain the fluctuation amplitude sequence. Compare the fluctuation amplitude sequence with a preset fluctuation threshold. If the threshold is exceeded, it is marked as abnormal.

[0029] If an anomaly is found, the wind speed distribution weight is adjusted based on the location of the anomaly point, and an iterative optimization algorithm is used to adjust the wind speed. In each iteration, the gradient descent method is used to minimize the fluctuation amplitude until the standard deviation of the fluctuation is lower than the preset threshold or the maximum number of iterations is reached, at which point the optimized airflow distribution pattern is output.

[0030] The time series wind speed values ​​are extracted based on the optimized airflow distribution pattern, and a continuous signal is generated using spline interpolation. The continuous signal is then low-pass filtered to obtain a smooth-transition wind speed control signal.

[0031] In conjunction with the first aspect, in the seventh implementation of the first aspect of this application, the ventilation equipment adjusts the wind speed intensity and ventilation frequency, including:

[0032] The wind speed control signal is converted into a digital command sequence and transmitted to the ventilation equipment actuator. The ventilation equipment actuator parses the digital command sequence to obtain the wind speed intensity value and ventilation frequency parameters.

[0033] In conjunction with the first aspect, in the eighth implementation of the first aspect of this application, determining the physiological demand vector adjustment value for the next cycle includes:

[0034] The air quality feedback data is input into the feather development model. The weight parameters of the feather development model are updated through machine learning algorithms. The weight gain rate curve and metabolic intensity fluctuation curve are recalculated based on the updated model. The growth slope and peak point are extracted from the weight gain rate curve, and the fluctuation amplitude and cycle length are extracted from the metabolic intensity fluctuation curve. The extracted parameters are mapped to the physiological demand vector space to generate the physiological demand vector adjustment value for the next cycle.

[0035] Secondly, this application provides an adaptive control system for poultry house environment, the system comprising:

[0036] The acquisition module collects poultry physiological data through a sensor array to obtain a weight gain rate curve and a metabolic intensity fluctuation curve. Based on the weight gain rate curve and the metabolic intensity fluctuation curve, and combined with a preset feather development model, it determines a physiological demand vector.

[0037] The assessment module evaluates the degree of feather development in poultry based on the physiological demand vector, obtains a quantitative score, analyzes the correlation between the quantitative score and the physiological demand vector, and calculates wind speed adjustment parameters based on the correlation.

[0038] The adjustment module generates an airflow distribution pattern sequence based on the wind speed adjustment parameters, determines whether there are any abnormal fluctuations, optimizes the wind speed control signal, and transmits the wind speed control signal to the ventilation equipment, which adjusts the wind speed intensity and ventilation frequency.

[0039] The feedback module collects real-time air quality feedback data in the poultry house, inputs the air quality feedback data into the feather development model, determines the physiological demand vector adjustment value for the next cycle, and generates a breeding benefit assessment report based on the adjustment value and dynamic fluctuation data.

[0040] Compared with the prior art, the beneficial effects of the present invention are at least as follows:

[0041] This invention provides an adaptive control method for poultry house environment. Based on physiological data such as poultry weight gain rate and metabolic intensity fluctuations, it adjusts environmental parameters, especially temperature and ventilation, in real time to optimize the internal environment of the poultry house. In this method, physiological data of poultry is collected through a sensor array and combined with a feather development model to accurately calculate the physiological demand vector of poultry. Based on these demands, wind speed and ventilation frequency are adjusted in real time, effectively reducing unnecessary energy consumption. Simultaneously, a fuzzy logic algorithm is used for wind speed adjustment, making the ventilation system more sensitive and precise, avoiding fluctuations and errors caused by fixed threshold adjustments in traditional methods, and significantly improving the stability and sustainability of ventilation effects. This invention can also dynamically adjust ventilation and temperature control parameters according to different growth stages of poultry and environmental changes, reducing the negative impact of stress on poultry growth, improving poultry comfort and health, and thus optimizing breeding efficiency. Furthermore, by collecting real-time air quality feedback data and inputting it into the feather development model, the physiological demand vector can be further adjusted, providing a more accurate basis for environmental control in the next cycle, effectively achieving a high degree of matching between environmental control and poultry needs, improving breeding efficiency, reducing energy consumption, and enhancing the overall health and production efficiency of poultry. Attached Figure Description

[0042] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a schematic diagram of an embodiment of an adaptive control method for poultry house environment in this application.

[0044] Figure 2 This is a flowchart of the fluctuation anomaly judgment process in the embodiments of this application;

[0045] Figure 3 This is a comparison diagram of ventilation effects in the embodiments of this application;

[0046] Figure 4 This is a schematic diagram of one embodiment of an adaptive control system for poultry house environment in this application. Detailed Implementation

[0047] This application provides a method and system for adaptive control of poultry house environment. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data used can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or device that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or devices.

[0048] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of the adaptive control method for poultry house environment in this application includes:

[0049] Step S1: Collect physiological data of poultry through sensor array to obtain weight gain rate curve and metabolic intensity fluctuation curve. Based on the weight gain rate curve and metabolic intensity fluctuation curve, determine the physiological demand vector in combination with the preset feather development model.

[0050] In modern poultry farming (such as chickens, ducks, geese, etc.), environmental factors have a significant impact on the growth and development of poultry. In particular, fluctuations in physiological indicators such as weight gain and metabolic intensity directly determine the health status and farming efficiency of poultry. Traditional methods of controlling the farming environment mostly rely on human experience and single sensor monitoring, lacking a precise understanding and dynamic adjustment of the physiological needs of poultry. To improve farming efficiency and reduce costs, this application uses a sensor array to collect poultry physiological data in real time and combines it with a preset feather development model to dynamically determine the physiological demand vector of poultry, thereby intelligently controlling environmental parameters. Specifically, multiple sensor arrays, including weighing sensors and infrared sensors, are deployed in the poultry house to collect poultry physiological data in real time, including weight data and metabolic index data (such as body temperature and respiratory rate). The collected weight data undergoes median filtering to remove outliers and noise, and then the weight increment per unit time is calculated to generate a weight gain rate curve. The metabolic index signals are low-pass filtered to remove high-frequency noise and interference, and the trend of metabolic intensity is extracted to generate a metabolic intensity fluctuation curve. The characteristic data of the weight gain rate curve and the metabolic intensity fluctuation curve are input into a preset feather development model. The feather development model outputs the analysis results of weight gain rate and metabolic intensity, generating a current physiological demand vector for the poultry. This vector includes environmental parameters such as temperature, humidity, and ventilation requirements. To facilitate subsequent environmental adjustments, the obtained physiological demand vector is standardized by using a min-max normalization method to map the values ​​of each demand item to the [0, 1] interval, ensuring data consistency and comparability. By dynamically generating the environmental parameters required for poultry growth based on real-time physiological data, the environmental control equipment can respond to the poultry's needs in real time, significantly improving the accuracy and dynamic adaptability of environmental regulation.

[0051] Step S2: Assess the degree of bird feather development based on physiological demand vectors, obtain quantitative scores, analyze the correlation between quantitative scores and physiological demand vectors, and calculate wind speed adjustment parameters based on the correlation.

[0052] Specifically, in poultry farming, feather development is strongly influenced by physiological needs (such as nutrition, temperature, and humidity). To precisely regulate the farming environment, this application proposes an assessment of poultry feather development based on a physiological demand vector. If a dimension of the physiological demand vector exceeds a preset threshold, a subsequent assessment process is triggered. Image acquisition devices are used to photograph poultry and obtain image data of their feathers. This image data is then input into a convolutional neural network algorithm to extract feather texture features. Based on these features, feather coverage density is calculated and mapped to a quantization score of 0-100. A fuzzy logic algorithm is used to analyze the correlation between the quantization score and the physiological demand vector. The correlation value is multiplied by a preset weighting factor to generate preliminary coefficients. Linear interpolation is then used to limit these preliminary coefficients within a preset range, yielding a wind speed intensity adjustment coefficient. Through the correlation analysis between the quantization score and the physiological demand vector, wind speed adjustment parameters are accurately calculated, optimizing air circulation and temperature and humidity control within the poultry house. This significantly improves the comfort of the farming environment, reduces the negative impact of stress on poultry growth, and enhances farming efficiency.

[0053] Step S3: Based on the wind speed adjustment parameters, generate an airflow distribution pattern sequence, determine whether there are any abnormal fluctuations, optimize the wind speed control signal, and transmit the wind speed control signal to the ventilation equipment, which then adjusts the wind speed intensity and ventilation frequency.

[0054] Specifically, in modern poultry farming, the air quality and airflow distribution within the poultry house directly affect the growth and health of the poultry. Wind speed and ventilation frequency are important parameters for regulating air circulation and maintaining suitable temperature and humidity in the poultry house. The wind speed adjustment parameter is input into the airflow distribution pattern generation function to generate a preliminary airflow distribution pattern sequence. This sequence represents the wind speed values ​​corresponding to a series of time points, ensuring that the wind speed covers the entire poultry house space, thereby achieving uniform air circulation. Fluctuation analysis is performed on the generated airflow distribution pattern sequence. By calculating the difference in wind speed values ​​at adjacent time points, a fluctuation amplitude sequence is obtained and compared with a preset fluctuation threshold. If the fluctuation... If the fluctuation exceeds a threshold, it is marked as abnormal, indicating uneven airflow distribution, which may cause localized high-temperature or low-oxygen areas, thus affecting poultry health. If abnormal fluctuations are detected, an iterative optimization algorithm is initiated to smooth wind speed changes by adjusting the airflow distribution pattern. Based on the optimized airflow distribution pattern, a continuous wind speed control signal is generated. The generated wind speed control signal is low-pass filtered to remove high-frequency noise, ensuring a smooth signal transition. The optimized wind speed control signal is stored in a database for later use. The wind speed control signal is transmitted to the actuator of the ventilation equipment via a wireless communication module, driving the ventilation equipment to adjust the wind speed intensity and ventilation frequency in real time. By dynamically adjusting the wind speed and ventilation frequency in the poultry house according to real-time physiological needs and environmental feedback, the stress response and environmental instability caused by excessive wind speed fluctuations are avoided. Furthermore, the optimized airflow distribution pattern not only improves the uniformity of air circulation in the poultry house but also ensures the stable operation of the ventilation equipment, thereby improving poultry comfort and growth efficiency and reducing the negative environmental impact on growth.

[0055] Step S4: Collect real-time air quality feedback data in the poultry house, input the air quality feedback data into the feather development model, determine the physiological demand vector adjustment value for the next cycle, and generate a breeding benefit assessment report based on the adjustment value and dynamic fluctuation data.

[0056] Specifically, air quality sensors installed at key locations in the poultry house collect real-time air quality data such as carbon dioxide concentration, ammonia level, temperature, and humidity. The collected air quality data is preprocessed to remove noise and irregular fluctuations, yielding air quality feedback data. This feedback data is then input into a feather development model, which is trained using historical growth data. By considering the impact of air quality changes on poultry growth, the model outputs a physiological demand vector for the current growth stage, including components for temperature, humidity, and ventilation requirements, thereby dynamically adjusting environmental control strategies. By analyzing trends in air quality data, such as increases or decreases in temperature and carbon dioxide concentration, combined with metabolic intensity fluctuation curves, it is determined whether adjustments to wind speed or ventilation frequency are necessary. Based on the adjusted physiological demand vector and dynamic fluctuation data, a poultry farming efficiency assessment report is generated. This report includes assessments of farming efficiency, energy consumption analysis, the effectiveness of temperature and humidity optimization, and poultry health status. Dynamic physiological demand adjustments based on real-time air quality data ensure that environmental regulation precisely adapts to the physiological changes of poultry, avoiding excessive or insufficient regulation. Furthermore, the combination of air quality feedback and feather development models effectively improves the comfort of the breeding environment, reduces stress responses, and thus improves poultry growth efficiency and breeding benefits. In addition, the generation of breeding benefit assessment reports can identify potential problems in a timely manner, further optimize environmental control strategies, and enhance the level of intelligence in breeding management.

[0057] In one specific embodiment, the process of collecting poultry physiological data through a sensor array to obtain a weight gain rate curve and a metabolic intensity fluctuation curve specifically includes the following steps:

[0058] The poultry physiological data collected by the sensor array includes poultry weight data and metabolic index signals. The poultry weight data and metabolic index signals are filtered to remove noise interference. Based on the filtered data, the weight gain rate is calculated and a weight gain rate curve is generated. Based on the filtered metabolic index signals, the changes in metabolic intensity are analyzed and a metabolic intensity fluctuation curve is generated.

[0059] Specifically, a weighing sensor is used to capture real-time changes in poultry weight, recording weight data every minute. Metabolic indicators, such as body temperature and respiratory rate, are collected using an infrared sensor. The collected weight and metabolic indicator signals are filtered to remove noise and interference. For weight data, a mid-range filter is applied with a window size of 5 sampling points to effectively remove sudden noise caused by poultry activity. For metabolic indicator data, a low-pass filter is used to suppress high-frequency interference while preserving the core frequency bands of the metabolic signal, ensuring signal stability. The filtered weight data is then differentially calculated to obtain the weight increment per unit time. A polynomial curve is fitted using the least squares method to generate a smooth weight gain rate curve. A Fourier transform is performed on the metabolic indicator signals to extract periodic fluctuation components and identify metabolic intensity fluctuation patterns. Based on the Fourier transform analysis results, spline interpolation is used to smoothly connect data points, generating a smooth metabolic intensity fluctuation curve, further quantifying the fluctuation amplitude. The generated weight gain rate curve and metabolic intensity fluctuation curve are converted into a standard format, and a timestamp and poultry identifier are added to each curve data. The data is then stored in the database to ensure data traceability. The stored weight gain rate curve and metabolic intensity fluctuation curve can be used as the basic data for the subsequent generation of physiological demand vectors. They can be input into the feather development model to further analyze the physiological needs of poultry and adjust the breeding environment in real time.

[0060] In one specific embodiment, obtaining the physiological demand vector for the current growth stage specifically includes the following steps:

[0061] By using a parameter fusion algorithm, feature data of the weight gain rate curve and metabolic intensity fluctuation curve are extracted. Based on historical poultry growth data and physiological index data, a feather development model is established and trained. The feature data is input into the feather development model, which divides the poultry growth process into multiple growth stages. Based on the threshold of the feature data, the current growth stage of the poultry is determined, and the physiological demand vector corresponding to the current growth stage is obtained, including nutritional demand, temperature demand and ventilation demand components.

[0062] Specifically, based on historical poultry growth data and physiological indicators, a feather development model was established using the Logistic growth function. This model describes the change in feather coverage over time, and the formula is as follows: ,in, For maximum feather coverage, For growth rate, To determine the inflection point time, historical poultry growth data and physiological indicators were fitted using the least squares method to determine the model parameters. A parameter fusion algorithm was employed to extract key feature data from the weight gain rate and metabolic intensity fluctuation curves. For example, a weighted average fusion method was used to sum the peak slope of the weight gain rate curve with the variance of the metabolic intensity fluctuation curve, and weights were calculated based on the Pearson correlation coefficient. Principal component analysis (PCA) was applied to the fused feature vector, retaining the first two principal components as the extracted feature data. This feature data was then input into the feather development model. Based on the threshold values ​​of the feature data, the feather development model determined the growth stage of the poultry, such as the initial growth stage, the growth stage, and the maturity stage. For example, when the average growth rate exceeded 0.04 kg / day, the poultry was classified as being in the growth stage, generating a corresponding physiological requirement vector, such as [25°C, 60% humidity, 0.5 m / s wind speed]. Through the above steps, based on the fluctuations in poultry weight gain rate and metabolic intensity, and combined with the feather development model, a dynamic physiological demand vector that meets the current growth stage can be generated. This not only provides a scientific basis for subsequent environmental control, but also ensures real-time optimization of the breeding environment, avoiding the negative impact of excessive or insufficient ventilation and temperature and humidity regulation on poultry growth.

[0063] In one specific embodiment, assessing the degree of feather development in birds based on physiological demand vectors to obtain a quantitative score specifically includes the following steps:

[0064] Determine whether the physiological demand vector exceeds a preset threshold. If it does, activate the dynamic evaluation module.

[0065] The dynamic evaluation module loads an image acquisition device to acquire images of poultry.

[0066] Feather texture features of bird images are extracted based on convolutional neural network algorithm, and feather coverage density is calculated;

[0067] A quantitative score for the degree of feather development is calculated based on the density of feather coverage.

[0068] Specifically, the values ​​of each dimension in the physiological demand vector, such as nutritional demand, temperature demand, and ventilation demand, are compared with preset thresholds. If any dimension exceeds the 80% threshold (for example, nutritional demand is greater than 80%), the entire physiological demand vector is determined to exceed the threshold, activating the dynamic evaluation module. This module loads an image acquisition device (such as a camera or infrared imager) to capture image data of poultry feathers. Using image recognition algorithms such as convolutional neural network algorithms, the acquired poultry images are first preprocessed to remove background noise, highlight feather areas, and extract feather texture features from the images, such as feather edges and color variations, to calculate feather coverage density. The specific calculation method will be explained later. Based on the feather coverage density, it is mapped to a quantization score of 0-100 using a weighting formula. Combined with historical data (such as the previous average feather coverage density), the quantization score is calculated using a weighted formula: Quantization score = density × 0.8 + historical average × 0.2. For example, if the current feather coverage density is 75% and the historical average is 70%, then the quantization score = 75 × 0.8 + 70 × 0.2 = 74. The density threshold and fraction calculation methods differ across different poultry farming scenarios (e.g., chicks, adult chickens). For example, for chicks, the density threshold is set at 50%, and an age factor is incorporated into the calculation. For adult chickens, the emphasis is on the uniformity of feather density. Score = Average Density × Uniformity Coefficient. By combining physiological demand vectors with image recognition algorithms, the degree of feather development in poultry is accurately assessed, and a quantitative score is generated. This allows for dynamic responses to changes in poultry growth stages, optimization of the breeding environment, and avoidance of the impact of excessive or insufficient environmental regulation on poultry.

[0069] In one specific embodiment, calculating the feather coverage density includes the following steps:

[0070] The feather region of a bird image is segmented at the pixel level using a semantic segmentation algorithm. The feather coverage area ratio is calculated based on the segmentation results to obtain the feather coverage density.

[0071] Specifically, a convolutional neural network model, particularly U-Net or Mask R-CNN in deep learning, is used for semantic segmentation. The model assigns a label to each pixel in the bird image, marking it as either a feather region or a background region. The semantically segmented image is then pixel-counted to calculate the number of pixels in the feather region. Assuming the total number of pixels in the image is... The number of pixels in the feather region is Feather cover density is calculated based on the following formula: Assuming the acquired bird images have a resolution of 1000x1000 pixels, =1,000,000 pixels, using a semantic segmentation algorithm, the number of pixels identified in the feather region is: If the image size is 700,000 pixels, then the feather coverage density of the bird is 70%, meaning that 70% of the image is covered by feathers. However, the color and texture of the feathers may be similar to the background color, causing segmentation errors. To improve calculation accuracy, a template matching method can be used to compare the current image with a standard feather template and adjust the deviation of the segmentation results.

[0072] In one specific embodiment, calculating wind speed adjustment parameters based on correlation specifically includes the following steps:

[0073] The fuzzy logic algorithm maps the quantified scores and physiological demand vectors to a preset fuzzy set, and combines the rule base to perform fuzzy inference to obtain the correlation value between poultry physiological demand and environmental regulation parameters.

[0074] The correlation value is multiplied by a preset weighting factor to obtain a preliminary coefficient, and the preliminary coefficient is limited to a preset range by linear interpolation to generate a wind speed intensity adjustment coefficient.

[0075] The air volume transition parameters are generated based on the wind speed intensity adjustment coefficient. The air volume transition parameters are generated in sequence from the current air volume to the target air volume.

[0076] Smooth the airflow transition parameters to obtain stable airflow adjustment parameters.

[0077] Specifically, the quantification score and the physiological demand vector are mapped to three fuzzy sets: "low", "medium", and "high". For example, if the quantification score is below 0.4, it is mapped to the "low" fuzzy set; if the quantification score is between 0.4 and 0.7, it is mapped to the "medium" fuzzy set; and if the quantification score is above 0.7, it is mapped to the "high" fuzzy set. Assuming that the temperature demand component in the physiological demand vector is 0.6 and the wind speed demand component is 0.8, then the temperature demand component 0.6 is mapped to the "medium" fuzzy set, and the wind speed demand component 0.8 is mapped to the "high" fuzzy set. Based on the parameter fusion of a bird feather development model, a fuzzy rule base was established. This rule base defines the correlation between quantification scores and physiological demand vectors. For example, if the quantification score is mapped to a "low" fuzzy set, and the wind speed demand component in the physiological demand vector is mapped to a "high" fuzzy set, the correlation is strong, indicating a need to increase ventilation to meet the air demand brought about by high metabolic intensity. If the quantification score is mapped to a "high" fuzzy set, and the temperature demand component in the physiological demand vector is mapped to a "medium" fuzzy set, the correlation is weak, indicating a lower temperature requirement due to environmental changes. Using fuzzy... Inference methods, such as the Mamdani method, employ fuzzy inference. This involves calculating the correlation between a quantified score and a physiological demand vector by comparing the input fuzzy set values ​​with rules in a rule base. For example, a quantified score of 0.3 (mapped to a "low" fuzzy set) and a wind speed demand of 0.8 (mapped to a "high" fuzzy set) indicate a strong correlation according to the rule base. Conversely, a quantified score of 0.5 (mapped to a "medium" fuzzy set) and a wind speed demand of 0.2 (mapped to a "low" fuzzy set) indicate a weaker correlation. The fuzzy inference result is then converted into a value between 0 and 1 using min-max operations, representing the closeness between physiological demand and environmental regulation. For instance, a correlation of 0.8 indicates a strong match between physiological demand and environmental regulation needs. If the quantified score is "medium" and the wind speed demand is "low," the correlation is 0.3. Based on the correlation value, a wind speed intensity adjustment coefficient is calculated. For example, a correlation value of 0.8 results in a higher adjustment coefficient, while a correlation value of 0.3 results in a lower coefficient. Finally, wind speed adjustment parameters are generated based on this coefficient to ensure that ventilation equipment adjusts wind speed to meet the physiological needs of poultry.

[0078] In one specific embodiment, generating an airflow distribution pattern sequence based on wind speed adjustment parameters and determining whether there are fluctuation anomalies specifically includes the following steps:

[0079] The wind speed adjustment parameters are input into the sequence generation function to calculate the initial airflow distribution pattern sequence;

[0080] Calculate the wind speed difference between adjacent time points in the initial airflow distribution pattern sequence to obtain the fluctuation amplitude sequence. Compare the fluctuation amplitude sequence with the preset fluctuation threshold. If it exceeds the threshold, it is marked as abnormal.

[0081] If an anomaly is found, the wind speed distribution weight is adjusted based on the location of the anomaly point, and an iterative optimization algorithm is used to adjust the wind speed. In each iteration, the gradient descent method is used to minimize the fluctuation amplitude until the standard deviation of the fluctuation is lower than the preset threshold or the maximum number of iterations is reached, at which point the optimized airflow distribution pattern is output.

[0082] The time series wind speed values ​​are extracted based on the optimized airflow distribution pattern, and a continuous signal is generated using spline interpolation. The continuous signal is then low-pass filtered to obtain a smooth-transitioning wind speed control signal.

[0083] Specifically, such as Figure 2 The diagram shows a flowchart for handling abnormal fluctuations. Wind speed adjustment parameters are input into an airflow distribution pattern generation function. This function calculates a series of wind speed distribution pattern sequences at different time points based on the input parameters. Assuming the wind speed adjustment parameters include a wind speed adjustment coefficient and ventilation frequency, and assuming the initial wind speed in the poultry house is 10 cubic meters per minute under basic conditions, the airflow distribution pattern generation function calculates the corresponding wind speed value at each time point using a wind speed adjustment coefficient of 1.5 and a ventilation frequency of 6 times per hour. Assuming the generated initial airflow distribution pattern is [10, 12, 14, 16, 15] (unit: cubic meters per minute), the difference in wind speed values ​​between adjacent time points in the initial airflow distribution pattern sequence is calculated to obtain a fluctuation amplitude sequence: [2, 2, 2, -1]. This fluctuation amplitude sequence is compared with a preset fluctuation threshold, assuming the preset threshold is 0.5 m / s. If the absolute value of the fluctuation amplitude is greater than this threshold, an abnormal fluctuation is considered to exist. In the above fluctuation amplitude sequence, the absolute values ​​of 2 and -1 both exceed 0.5, therefore they are marked as abnormal. If abnormal fluctuations are detected, the weights in the wind speed distribution pattern are adjusted based on the location of the abnormal point. For example, if a sudden increase or decrease in wind speed is detected at a certain moment (such as the third time point), the system will adjust the wind speed weights near that point to make the wind speed changes more stable. During the wind speed adjustment process, an iterative optimization algorithm (such as gradient descent) is applied to minimize the fluctuation amplitude. By gradually updating the wind speed distribution pattern, the fluctuation amplitude is ensured to be reduced below the threshold. In each iteration, the parameters of the wind speed distribution are adjusted by calculating the gradient of the loss function to reduce the fluctuation amplitude and obtain an optimized airflow distribution pattern. Wind speed values ​​are extracted from the optimized airflow distribution pattern, and a continuous wind speed control signal is generated using spline interpolation. Spline interpolation can connect discrete wind speed points by fitting a smooth curve to ensure a smooth transition of the wind speed control signal. By combining fuzzy logic algorithms with optimization algorithms, the wind speed distribution can be corrected through iterative optimization algorithms when abnormal wind speed fluctuations occur, making wind speed adjustments more accurate and ensuring the comfort of poultry at different growth stages.

[0084] In one specific embodiment, adjusting the wind speed and ventilation frequency of the ventilation equipment specifically includes the following steps:

[0085] The wind speed control signal is converted into a digital command sequence and transmitted to the ventilation equipment actuator. The ventilation equipment actuator parses the digital command sequence to obtain the wind speed intensity value and ventilation frequency parameters.

[0086] Specifically, a preset encoding protocol is used to convert wind speed control signals into digital command sequences. The encoding protocol includes converting parameters such as wind speed and ventilation frequency into digital signals using ASCII or binary encoding, followed by necessary verification and encryption. Assuming a wind speed of 4 m / s and a ventilation frequency of 6 times per hour, the generated digital command sequence is "V4.0T6.0", where "V" represents wind speed and "T" represents ventilation frequency. This digital command sequence is sent to the actuator interface via a wireless communication module (such as Wi-Fi, Bluetooth, or Zigbee). In practical applications, multiple control signals may be transmitted simultaneously, requiring priority sequencing. Typically, the priority of wind speed adjustment signals depends on the urgency of the adjustment. Commands requiring significant wind speed adjustments (e.g., from 2 m / s to 4 m / s) have higher priority. This priority sequencing mechanism ensures that wind speed control signals are efficiently transmitted to the actuator and begin adjustment in the shortest possible time, avoiding the impact of transmission delays on the real-time performance of poultry house ventilation. After receiving the digital command sequence, the actuator first analyzes and extracts the values ​​of wind speed and ventilation frequency. For example, from the command "V4.0T6.0", the wind speed is extracted to be 4 m / s and the ventilation frequency is 6 times / hour. Based on the analyzed parameters, the actuator drives the motor to adjust the fan speed. The wind speed is typically set between 2 and 5 m / s, while the ventilation frequency is set to 4 to 8 times / hour based on the poultry's growth needs. If the wind speed is adjusted to 4 m / s, the actuator will control the motor speed to achieve that speed. The fan speed and ventilation frequency need to be precisely adjusted according to the poultry's current growth stage and physiological needs. Through this method, the wind speed control signal can be accurately and in real-time transmitted to the ventilation equipment actuator, and the actuator can precisely adjust the fan speed and wind speed based on the analyzed wind speed and ventilation frequency. The priority sorting mechanism ensures that the signal can be responded to quickly, thereby optimizing the air quality in the poultry house.

[0087] In one specific embodiment, determining the physiological demand vector adjustment value for the next cycle specifically includes the following steps:

[0088] Air quality feedback data is input into the feather development model. The weight parameters of the feather development model are updated through machine learning algorithms. The weight gain rate curve and metabolic intensity fluctuation curve are recalculated based on the updated model. The growth slope and peak point are extracted from the weight gain rate curve, and the fluctuation amplitude and cycle length are extracted from the metabolic intensity fluctuation curve. The extracted parameters are mapped to the physiological demand vector space to generate the physiological demand vector adjustment value for the next cycle.

[0089] Specifically, after adjusting the wind speed and ventilation frequency, the ventilation equipment collects real-time air quality feedback data. Assuming the air quality feedback data includes a carbon dioxide concentration of 650 ppm, an ammonia concentration of 35 ppm, a temperature of 28°C, and a humidity of 60%, the collected air quality feedback data is input into a feather development model. This model is trained based on historical weight data and physiological indicators, and combines environmental data to predict the physiological needs of poultry. The model's weight parameters are updated using machine learning algorithms. Based on the current growth stage's weight data, the model calculates the weight gain rate curve for the next cycle. Assuming the current growth stage's weight gain rate is 0.05 kg per day, the updated prediction model may predict an increase in the growth rate to 0.07 kg per day based on environmental adjustment needs. Based on the same update... The updated model calculates the metabolic intensity fluctuation curve. Assuming the current metabolic intensity fluctuation range is 20%, the updated model may show a new periodic fluctuation in the metabolic intensity fluctuation curve, reflecting the impact of environmental changes on metabolic intensity. The growth slope is extracted from the weight gain rate curve, representing the poultry's weight gain rate at the current stage. When extracting the fluctuation range and period length from the metabolic intensity fluctuation curve, the extracted parameters (such as growth slope and fluctuation range) are mapped to the physiological demand vector space to generate the physiological demand vector adjustment value for the next period. For example, if the weight gain rate slope is 0.07 kg / day and the metabolic intensity fluctuation range is 25%, the generated physiological demand vector adjustment value might be: nutritional demand [10% increase], temperature demand [unchanged], and ventilation demand [15% increase]. Based on real-time collected air quality feedback data and machine learning algorithms, the weight parameters of the feather development model are updated, enabling accurate calculation of the poultry's physiological demand vector adjustment value. Through this process, the system can dynamically adapt to environmental changes, ensuring that environmental regulation parameters such as wind speed, temperature, and ventilation frequency are precisely matched to the actual needs of poultry at different growth stages, reducing unnecessary resource waste, improving poultry comfort, and promoting healthy growth.

[0090] like Figure 3The diagram shows a comparison of ventilation effects. Traditional ventilation methods (solid gray line): This line illustrates the changes in ventilation intensity within the poultry house using traditional methods. Traditional methods typically adjust ventilation based on fixed thresholds, which may lack flexibility and result in significant fluctuations in ventilation intensity. This reflects the slow response of traditional methods to environmental changes and the potential for over-ventilation or under-ventilation. Ventilation using the method described in this application (dashed black line): This line illustrates the ventilation adjustment based on the method described in this application. This method adjusts ventilation intensity in real time based on data such as poultry physiological needs, feather development, and metabolic intensity. It allows for more precise control of ventilation intensity, making it more stable and responsive to environmental changes. Through this adaptive regulation, ventilation intensity fluctuations are relatively small, and it is better maintained within a suitable range.

[0091] The above describes an adaptive control method for poultry house environment in an embodiment of this application. The following describes an adaptive control system for poultry house environment in an embodiment of this application. Please refer to [link / reference]. Figure 4 One embodiment of the adaptive control system for poultry house environment in this application includes:

[0092] The data acquisition module collects physiological data of poultry through a sensor array, obtains the weight gain rate curve and metabolic intensity fluctuation curve, and determines the physiological demand vector based on the weight gain rate curve and metabolic intensity fluctuation curve, combined with a preset feather development model.

[0093] The assessment module evaluates the degree of feather development in poultry based on physiological demand vectors, obtains a quantitative score, analyzes the correlation between the quantitative score and the physiological demand vector, and calculates wind speed adjustment parameters based on the correlation.

[0094] The adjustment module generates an airflow distribution pattern sequence based on wind speed adjustment parameters, determines whether there are any abnormal fluctuations, optimizes the wind speed control signal, and transmits the wind speed control signal to the ventilation equipment, which then adjusts the wind speed intensity and ventilation frequency.

[0095] The feedback module collects real-time air quality feedback data in the poultry house, inputs the air quality feedback data into the feather development model, determines the physiological demand vector adjustment value for the next cycle, and generates a breeding benefit assessment report based on the adjustment value and dynamic fluctuation data.

[0096] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0097] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0098] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for adaptive control of poultry house environment, characterized in that, The method includes: Step S1: Collect poultry physiological data through a sensor array to obtain a weight gain rate curve and a metabolic intensity fluctuation curve. Based on the weight gain rate curve and the metabolic intensity fluctuation curve, determine the physiological demand vector in conjunction with a preset feather development model. Step S2: Evaluate the degree of bird feather development based on the physiological demand vector to obtain a quantitative score, analyze the correlation between the quantitative score and the physiological demand vector, and calculate the wind speed adjustment parameters based on the correlation; Step S3: Based on the wind speed adjustment parameters, generate an airflow distribution pattern sequence, determine whether there are any abnormal fluctuations, optimize the wind speed control signal, and transmit the wind speed control signal to the ventilation equipment, which adjusts the wind speed intensity and ventilation frequency. Step S4: Collect air quality feedback data in the poultry house in real time, input the air quality feedback data into the feather development model, determine the physiological demand vector adjustment value for the next cycle, and generate a breeding benefit assessment report based on the adjustment value and dynamic fluctuation data.

2. The method according to claim 1, characterized in that, Physiological data of poultry were collected using a sensor array, resulting in curves showing the rate of weight gain and metabolic intensity fluctuations, including: The poultry physiological data collected by the sensor array includes poultry weight data and metabolic index signals. The poultry weight data and metabolic index signals are processed by signal filtering to remove noise interference. Based on the filtered data, the weight gain rate is calculated and a weight gain rate curve is generated. Based on the filtered metabolic index signals, the metabolic intensity change is analyzed and a metabolic intensity fluctuation curve is generated.

3. The method according to claim 1, characterized in that, Obtaining the physiological demand vector for the current growth stage includes: The feature data of the weight gain rate curve and the metabolic intensity fluctuation curve are extracted by the parameter fusion algorithm. A feather development model is established and trained based on historical poultry growth data and physiological index data. The feature data is input into the feather development model. The feather development model divides the growth process of poultry into multiple growth stages. The current growth stage of poultry is determined based on the threshold of the feature data. The physiological demand vector corresponding to the current growth stage is obtained, including nutritional demand, temperature demand and ventilation demand components.

4. The method according to claim 1, characterized in that, The degree of feather development in birds is assessed based on the aforementioned physiological demand vector, resulting in a quantitative score including: Determine whether the physiological demand vector exceeds a preset threshold. If it does, activate the dynamic evaluation module. The dynamic evaluation module loads an image acquisition device to acquire images of poultry; The feather texture features of the bird images are extracted using a convolutional neural network algorithm, and the feather coverage density is calculated using the following formula: %,in, This represents the total number of pixels in the bird image. The number of pixels in the feather region. Feather coverage density; A quantitative score for the degree of feather development is calculated based on the stated feather coverage density.

5. The method according to claim 4, characterized in that, Calculating feather cover density includes: The feather region of a bird image is segmented at the pixel level using a semantic segmentation algorithm. The feather coverage area ratio is calculated based on the segmentation results to obtain the feather coverage density.

6. The method according to claim 1, characterized in that, The wind speed adjustment parameters calculated based on correlation include: The quantified score and the physiological requirement vector are mapped to a preset fuzzy set using a fuzzy logic algorithm. Fuzzy reasoning is then performed using a rule base to obtain the correlation value between poultry physiological requirements and environmental regulation parameters. The correlation value is multiplied by a preset weighting factor to obtain a preliminary coefficient, and the preliminary coefficient is limited to a preset range by linear interpolation to generate a wind speed intensity adjustment coefficient. The air volume transition parameters are generated based on the wind speed intensity adjustment coefficient, wherein the air volume transition parameters gradually transition from the current air volume to the target air volume in a sequence. The airflow transition parameters are smoothed to obtain stable airflow adjustment parameters.

7. The method according to claim 1, characterized in that, Based on the wind speed adjustment parameters, a sequence of airflow distribution patterns is generated, and the presence of abnormal fluctuations is determined, including: The wind speed adjustment parameters are input into the sequence generation function to calculate the initial airflow distribution pattern sequence; Calculate the wind speed difference between adjacent time points in the initial airflow distribution pattern sequence to obtain the fluctuation amplitude sequence. Compare the fluctuation amplitude sequence with a preset fluctuation threshold. If the threshold is exceeded, it is marked as abnormal. If an anomaly is found, the wind speed distribution weight is adjusted based on the location of the anomaly point, and an iterative optimization algorithm is used to adjust the wind speed. In each iteration, the gradient descent method is used to minimize the fluctuation amplitude until the standard deviation of the fluctuation is lower than the preset threshold or the maximum number of iterations is reached, at which point the optimized airflow distribution pattern is output. The time series wind speed values ​​are extracted based on the optimized airflow distribution pattern, and a continuous signal is generated using spline interpolation. The continuous signal is then low-pass filtered to obtain a smooth-transition wind speed control signal.

8. The method according to claim 1, characterized in that, The ventilation equipment adjusts the wind speed intensity and ventilation frequency, including: The wind speed control signal is converted into a digital command sequence and transmitted to the ventilation equipment actuator. The ventilation equipment actuator parses the digital command sequence to obtain the wind speed intensity value and ventilation frequency parameters.

9. The method according to claim 1, characterized in that, Determining the physiological demand vector adjustment values ​​for the next cycle includes: The air quality feedback data is input into the feather development model. The weight parameters of the feather development model are updated through machine learning algorithms. The weight gain rate curve and metabolic intensity fluctuation curve are recalculated based on the updated model. The growth slope and peak point are extracted from the weight gain rate curve, and the fluctuation amplitude and cycle length are extracted from the metabolic intensity fluctuation curve. The extracted parameters are mapped to the physiological demand vector space to generate the physiological demand vector adjustment value for the next cycle.

10. A poultry house environment adaptive control system, used to implement the poultry house environment adaptive control method as described in any one of claims 1-9, characterized in that, The system includes: The acquisition module collects poultry physiological data through a sensor array to obtain a weight gain rate curve and a metabolic intensity fluctuation curve. Based on the weight gain rate curve and the metabolic intensity fluctuation curve, and combined with a preset feather development model, it determines a physiological demand vector. The assessment module evaluates the degree of feather development in poultry based on the physiological demand vector, obtains a quantitative score, analyzes the correlation between the quantitative score and the physiological demand vector, and calculates wind speed adjustment parameters based on the correlation. The adjustment module generates an airflow distribution pattern sequence based on the wind speed adjustment parameters, determines whether there are any abnormal fluctuations, optimizes the wind speed control signal, and transmits the wind speed control signal to the ventilation equipment, which adjusts the wind speed intensity and ventilation frequency. The feedback module collects real-time air quality feedback data in the poultry house, inputs the air quality feedback data into the feather development model, determines the physiological demand vector adjustment value for the next cycle, and generates a breeding benefit assessment report based on the adjustment value and dynamic fluctuation data.

Images