Modernized feed supply control system and method for broiler farm

Abstract

The application discloses a modernized feed supply control system and method for broiler farms, and relates to the technical field of broiler farming automation. The system comprises a multidimensional data acquisition unit, a limited feeding influence evaluation unit and a dynamic feeding correction unit. First, multidimensional data of the chicken group is acquired. Second, multidimensional normal distribution modeling and analysis are performed on the multidimensional data of the chicken group, group state distribution data is extracted, and the group feeding vitality index is obtained by analyzing the feeding sound data. The group state distribution data and the group feeding vitality index are evaluated to generate a limited feeding influence evaluation result. Finally, whether the current limited feeding scheme needs to be corrected is determined according to the limited feeding influence evaluation result. When correction is needed, the feeding scheme is determined according to the group state distribution data and is executed. The application realizes accurate quantification and dynamic correction of the influence of the limited feeding scheme on the chicken group through multidirectional electrical impedance analysis and normal distribution form evaluation.

Description

Technical Field

[0001] This invention relates to the field of automation technology in broiler farming, specifically to a modern feed supply control system and method for broiler farms. Background Technology

[0002] Different feeding strategies are required at different growth stages of broilers. For example, during the growth period, feed restriction is required to control the rate of weight gain, improve flock uniformity, and reduce the incidence of leg disease and ascites. Therefore, a modern feed supply control system and method for broiler farms are needed.

[0003] In existing technologies, adjustments to feed restriction programs primarily rely on the experience and judgment of farm workers. Farm workers periodically and manually select a portion of the chickens from the flock for weighing, calculate the average weight of the sampled chickens, and compare it to a standard weight curve. If the average weight is lower than the standard value, the feed amount is increased; if it is higher than the standard value, the feed amount is decreased. Some farms with a higher degree of automation have installed automatic weighing platforms in the chicken house passageways. The weight data is automatically recorded as the chickens pass through the weighing platform during their daily activities. The system automatically calculates the average weight of the entire flock and automatically adjusts the feed amount according to preset threshold rules.

[0004] The above-mentioned solutions have the following technical problems: 1. Existing technologies rely solely on weight as a single dimension for adjusting feed restriction programs, failing to obtain information closely related to metabolism and health, such as body water content and tissue health status. When chickens experience dehydration or edema, weight data may fluctuate misleadingly, making it difficult to distinguish between an inappropriate feed restriction program and a health problem based solely on weight changes.

[0005] 2. Current technologies for analyzing weight data are limited to comparing mean values, focusing only on whether the average weight of the entire flock deviates from the standard value, while neglecting the analysis of weight distribution patterns. When a flock exhibits differentiation, i.e., some chickens are underweight and some are overweight, but the average weight remains within the standard range, the system cannot identify this abnormal differentiation, leading to the inability to adjust the feed restriction program in a timely manner and a continuous deterioration in flock uniformity.

[0006] 3. Existing technologies for collecting electrical impedance data typically employ a unidirectional current injection method, which can only obtain electrical impedance values ​​in one direction. This fails to reflect the differences in conductivity characteristics of chicken foot tissue in different directions, resulting in insufficient information dimensions and making it difficult to comprehensively assess the tissue health status of chickens.

[0007] 4. Existing technologies lack the ability to perceive the vitality of chicken flocks' feeding behavior and cannot determine whether the feed restriction program has caused excessive hunger stress to the flock. When the feed restriction intensity is too high, the flock's feeding enthusiasm will increase abnormally, manifested as scrambling for food and dense pecking sounds, but insufficient energy per pecking. This abnormal behavior cannot be identified by simply adjusting the amount of feed.

[0008] 5. The evaluation results and corrective decisions of existing technologies are mixed together. The results of weight assessment, health assessment and behavior assessment are not generated separately. When problems occur in the feed restriction program, it is difficult to trace the specific source of the problem and it is impossible to take targeted corrective measures in different dimensions such as adjusting the total amount of feed, differentiating feeding in different areas, adjusting nutrient concentration, and regulating drinking water. Summary of the Invention

[0009] In view of the above-mentioned technical deficiencies, the purpose of this invention is to provide a modern feed supply control system and method for broiler farms.

[0010] To solve the above-mentioned technical problems, the present invention adopts the following technical solution: The present invention provides a feed supply control system for modern broiler farms, including the following units: a multi-dimensional data acquisition unit, used to collect multi-dimensional data of the flock and perform standardized processing, the multi-dimensional data of the flock includes group weight distribution data, individual bioelectrical impedance data and feeding sound data.

[0011] The feed restriction impact assessment unit is used to perform multidimensional normal distribution modeling on population weight distribution data and individual bioelectrical impedance data, extract population state distribution data, analyze feeding sound data to obtain the overall feeding activity index, assess the weight distribution state and bioelectrical impedance distribution state of the population state distribution data respectively, assess the feeding activity state of the overall feeding activity index, and generate the feed restriction impact assessment results.

[0012] The dynamic feeding correction unit is used to determine whether the current feeding restriction plan needs to be corrected based on the assessment results of the impact of feeding restriction. When correction is needed, the feeding plan is determined based on the population status distribution data and then issued for execution.

[0013] Preferably, the multidimensional normal distribution modeling of the group weight distribution data and individual bioelectrical impedance data is carried out as follows: the group weight distribution data includes the weight value of each chicken, and the individual bioelectrical impedance data includes the multidirectional impedance data of each chicken.

[0014] The weight values ​​of each chicken are counted according to the preset weight range. A weight frequency distribution histogram is constructed with the weight value on the horizontal axis and the number of chickens in each weight range on the vertical axis. The weight frequency distribution histogram is then fitted with a Gaussian function to obtain the normal distribution curve of the weight.

[0015] The multi-directional impedance data of each chicken are grouped according to the current injection direction. The impedance value of each directional channel is counted according to the preset impedance interval. The impedance value is plotted on the horizontal axis and the number of chickens in each impedance interval is plotted on the vertical axis to construct the impedance frequency distribution histogram of each directional channel. The histogram of impedance frequency distribution of each directional channel is fitted with a Gaussian function to obtain the normal distribution curve of impedance of each directional channel. The normal distribution curve of body weight and the normal distribution curve of channel impedance in each direction are combined into a multidimensional normal distribution model.

[0016] Preferably, the multi-directional impedance data is acquired through the following process: the data is collected by a partitioned electrode array laid on the ground of the chicken passageway. The partitioned electrode array is divided into a front zone and a rear zone along the length of the passageway. The front zone is divided into a left front electrode and a right front electrode along the width of the passageway, and the rear zone is divided into a left rear electrode and a right rear electrode along the width of the passageway.

[0017] When the chicken walks through the partitioned electrode array naturally, the chicken's left foot steps on the left front electrode and the left rear electrode, and the chicken's right foot steps on the right front electrode and the right rear electrode. The partitioned electrode array injects current between each pair of electrodes in a preset sequence and measures the voltage response. The current injection direction includes a first longitudinal channel along the front and back direction of the chicken's left foot, a second longitudinal channel along the front and back direction of the chicken's right foot, and a transverse channel between the two feet of the chicken.

[0018] The first longitudinal channel is composed of the left front electrode and the left rear electrode, the second longitudinal channel is composed of the right front electrode and the right rear electrode, and the transverse channel is composed of the electrode contacted by the left foot and the electrode contacted by the right foot. The impedance values ​​measured by each channel are used as the multi-directional impedance data of the chicken.

[0019] Preferably, the extraction of population state distribution data is specifically performed as follows: the mean parameter and standard deviation parameter are extracted from the normal distribution curve of weight, the mean parameter is used as the weight peak position, the current weight peak position is compared with the standard weight peak position of the current age, and the weight peak offset is calculated.

[0020] The mean and standard deviation parameters of the impedance normal distribution curves of each directional channel are extracted. The mean parameter of each directional channel is used as the impedance peak position of that directional channel. The current impedance peak position of each directional channel is compared with the impedance peak position of the same directional channel in the previous feeding cycle, and the impedance peak offset of each directional channel is calculated.

[0021] Take a preset unit distance point to the left and right of the peak position of the normal distribution curve of weight. The preset unit distance is the distance of the standard deviation of the peak position offset to both sides by a preset multiple. Calculate the slope of the line connecting the preset unit distance point on the left to the peak position as the left slope value, and calculate the slope of the line connecting the preset unit distance point on the right to the peak position as the right slope value. Divide the left slope value by the absolute value of the right slope value to obtain the slope ratio of the two sides of the weight.

[0022] For the normal distribution curves of impedance in each direction, extract the left slope value, right slope value, and the ratio of the slopes on both sides of each direction channel in the same way.

[0023] The ratio of the first longitudinal channel impedance value to the second longitudinal channel impedance value is used as the longitudinal symmetry index, and the ratio of the transverse channel impedance value to the average of the first and second longitudinal channel impedance values ​​is used as the transverse conduction index.

[0024] The peak weight position, peak weight offset, slope ratio on both sides of the weight, peak impedance position in each direction, peak impedance offset in each direction, slope ratio on both sides of each direction, longitudinal symmetry index, and transverse conduction index are summarized into population state distribution data.

[0025] On the other hand, the present invention provides a modern feed supply control method for broiler farms, including the following process: Process 1, multidimensional data acquisition: collect multidimensional data of the flock and perform standardized processing. The multidimensional data of the flock includes group weight distribution data, individual bioelectrical impedance data and feeding sound data.

[0026] Step 2, Feed Restriction Impact Assessment: Multidimensional normal distribution modeling is performed on the group weight distribution data and individual bioelectrical impedance data to extract the group state distribution data. Feeding sound data is analyzed to obtain the overall group feeding activity index. The group state distribution data is assessed for weight distribution and bioelectrical impedance distribution, and the overall group feeding activity index is assessed for feeding activity status, generating the feed restriction impact assessment results.

[0027] Step 3: Dynamic Feeding Adjustment: Based on the impact assessment results of feed restriction, determine whether the current feed restriction plan needs to be adjusted. If adjustment is required, determine the feeding plan based on the population status distribution data and issue it for implementation.

[0028] The beneficial effects of the present invention are as follows: 1. By introducing multi-directional impedance data acquisition, the present invention uses a partitioned electrode array to obtain multi-channel impedance values ​​along the longitudinal and transverse directions of the foot during the natural walking process of chickens, thereby achieving multi-dimensional non-destructive detection of the tissue state of chickens without interrupting their normal activities, overcoming the limitations of traditional methods that rely solely on weight data.

[0029] 2. This invention constructs a multidimensional normal distribution model of body weight and multidirectional electrical impedance, and extracts distribution morphology characteristic parameters such as peak position, slope ratio on both sides, longitudinal symmetry index and transverse conduction index. This improves the assessment of flock status from simple mean comparison to geometric analysis of distribution morphology, enabling earlier and more sensitive detection of abnormal trends such as flock differentiation, dehydration or edema.

[0030] 3. This invention breaks down the evaluation results into three independent dimensions: body weight distribution status evaluation, electrical impedance distribution status evaluation, and feed intake activity status evaluation, and generates corresponding sub-evaluation results for each. This allows the source of problems in the feed restriction program to be accurately located, avoiding the difficulty of tracing the source of problems caused by mixed evaluation of multiple dimensions.

[0031] 4. This invention determines specific correction operations based on the results of sub-item evaluations, including the combined application of various correction methods such as adjusting the total amount of feed, differentiated feeding by region, regulating drinking water, supplementing additives, and adjusting feeding frequency, thereby achieving refined dynamic correction of the feed restriction program. Attached Figure Description

[0032] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0033] Figure 1 This is a schematic diagram of the system structure connection of the present invention.

[0034] Figure 2 This is a schematic diagram illustrating the implementation process of the method of the present invention. Detailed Implementation

[0035] 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.

[0036] according to Figure 1 As shown, the present invention provides a modern feed supply control system for broiler farms, comprising the following units: a multi-dimensional data acquisition unit, a feed restriction impact assessment unit, and a feed restriction impact assessment unit.

[0037] The feed restriction impact assessment unit is connected to both the multidimensional data acquisition unit and the feed restriction impact assessment unit.

[0038] The multidimensional data acquisition unit is used to collect multidimensional data of the chicken flock and perform standardized processing. The multidimensional data of the chicken flock includes group weight distribution data, individual bioelectrical impedance data and feeding sound data.

[0039] In one specific embodiment, the collection and standardization of multidimensional data of the chicken flock is carried out as follows: The collection of multidimensional data of the chicken flock is carried out as follows: the weight distribution data of the flock is collected by a weighing platform laid on the ground of the chicken passage. When the chickens walk through the weighing platform, the weighing sensor collects the weight value of the chickens, and the weight value is linked with the collection time and then uploaded.

[0040] It should be noted that the weighing platform is installed in the chicken house on the passageway that chickens must pass through to and from the drinking area and the feeding area. The chickens naturally pass through the weighing platform during their daily activities, without the need for manual herding or catching, thus reducing stress on the chickens.

[0041] The weighing platform is covered with anti-slip material to ensure that the chickens can pass through smoothly.

[0042] Each chicken has a unique identification tag stored in its electronic leg band. An RFID reader next to the weighing platform automatically reads the identification tag as the chicken passes by and binds it to its weight.

[0043] Individual bioelectrical impedance data were collected using a partitioned electrode array laid on the ground in the chicken passageway.

[0044] The sound data of feeding is collected by microphones deployed above each feed tray. The microphones collect audio signals from the feed tray area at a preset sampling frequency, and then upload the audio signals after binding them with the feed tray number and the collection time.

[0045] The collected group weight distribution data, individual bioelectrical impedance data, and feeding sound data are uniformly converted into a preset standard data format.

[0046] It should be noted that the standard data format includes a data source field, a collection timestamp field, a data type field, a numeric field, and a unit field.

[0047] The feed restriction impact assessment unit is used to perform multidimensional normal distribution modeling on population weight distribution data and individual bioelectrical impedance data, extract population state distribution data, analyze feeding sound data to obtain the overall feeding activity index, assess the weight distribution state and bioelectrical impedance distribution state of the population state distribution data respectively, assess the feeding activity state of the overall feeding activity index, and generate the feed restriction impact assessment results.

[0048] In one specific embodiment, the multidimensional normal distribution modeling of the group weight distribution data and individual bioelectrical impedance data is performed as follows: the group weight distribution data includes the weight value of each chicken, and the individual bioelectrical impedance data includes the multidirectional impedance data of each chicken.

[0049] The weight values ​​of each chicken are counted according to the preset weight range. A weight frequency distribution histogram is constructed with the weight value on the horizontal axis and the number of chickens in each weight range on the vertical axis. The weight frequency distribution histogram is then fitted with a Gaussian function to obtain the normal distribution curve of the weight.

[0050] It should be noted that the preset weight range is determined based on the standard weight range of the broiler breed and age, and the range width is set as a preset percentage of the standard weight.

[0051] The Gaussian function fitting uses the least squares method, and the mean and standard deviation parameters are obtained by iteratively solving to minimize the sum of squared residuals between the Gaussian function and the frequency distribution histogram.

[0052] The multi-directional impedance data of each chicken are grouped according to the current injection direction. The impedance value of each directional channel is counted according to the preset impedance interval. The impedance value is plotted on the horizontal axis and the number of chickens in each impedance interval is plotted on the vertical axis to construct the impedance frequency distribution histogram of each directional channel. The histogram of impedance frequency distribution of each directional channel is fitted with a Gaussian function to obtain the normal distribution curve of impedance of each directional channel.

[0053] It should be noted that the collected individual electrical impedance data are not converted into body fat percentage or lean body mass. Instead, the original electrical impedance values ​​are used directly to model a normal distribution, thus avoiding the cumulative error introduced by the intermediate conversion process.

[0054] The normal distribution curve of body weight and the normal distribution curve of channel impedance in each direction are combined into a multidimensional normal distribution model.

[0055] In one specific embodiment, the multi-directional impedance data is acquired through the following process: the data is acquired by a partitioned electrode array laid on the ground of the chicken passageway. The partitioned electrode array is divided into a front zone and a rear zone along the length of the passageway. The front zone is divided into a left front electrode and a right front electrode along the width of the passageway, and the rear zone is divided into a left rear electrode and a right rear electrode along the width of the passageway.

[0056] It should be noted that the area of ​​each electrode in the partitioned electrode array is a preset value. The electrode surface is flush with the channel floor and is treated with anti-slip material. There are insulating isolation strips between the left front electrode, right front electrode, left rear electrode and right rear electrode to ensure electrical insulation between each electrode.

[0057] When the chicken walks through the partitioned electrode array naturally, the chicken's left foot steps on the left front electrode and the left rear electrode, and the chicken's right foot steps on the right front electrode and the right rear electrode. The partitioned electrode array injects current between each pair of electrodes in a preset sequence and measures the voltage response. The current injection direction includes a first longitudinal channel along the front and back direction of the chicken's left foot, a second longitudinal channel along the front and back direction of the chicken's right foot, and a transverse channel between the two feet of the chicken.

[0058] It should be noted that the preset timing is to first perform current injection and voltage measurement of the first vertical channel, then perform current injection and voltage measurement of the second vertical channel, and finally perform current injection and voltage measurement of the horizontal channel. The measurement time of each channel is within the preset time range.

[0059] The injected current is a weak AC signal of a preset frequency, and the current intensity is controlled within a preset safe range so as not to cause any stimulation or discomfort to the chickens.

[0060] The voltage response is acquired by a differential amplifier circuit, and the impedance value is calculated after analog-to-digital conversion.

[0061] The first longitudinal channel is composed of the left front electrode and the left rear electrode, the second longitudinal channel is composed of the right front electrode and the right rear electrode, and the transverse channel is composed of the electrode contacted by the left foot and the electrode contacted by the right foot. The impedance values ​​measured by each channel are used as the multi-directional impedance data of the chicken.

[0062] In one specific embodiment, the extraction of population state distribution data is carried out as follows: the mean parameter and standard deviation parameter are extracted from the normal distribution curve of weight, the mean parameter is used as the weight peak position, the current weight peak position is compared with the standard weight peak position of the current age, and the weight peak offset is calculated.

[0063] It should be noted that the peak weight offset is the difference between the current peak weight position and the standard peak weight position. A positive value indicates that the group is overweight, and a negative value indicates that the group is underweight.

[0064] The mean and standard deviation parameters of the impedance normal distribution curves of each directional channel are extracted. The mean parameter of each directional channel is used as the impedance peak position of that directional channel. The current impedance peak position of each directional channel is compared with the impedance peak position of the same directional channel in the previous feeding cycle, and the impedance peak offset of each directional channel is calculated.

[0065] Take a preset unit distance point to the left and right of the peak position of the normal distribution curve of weight. The preset unit distance is the distance of the standard deviation of the peak position offset to both sides by a preset multiple. Calculate the slope of the line connecting the preset unit distance point on the left to the peak position as the left slope value, and calculate the slope of the line connecting the preset unit distance point on the right to the peak position as the right slope value. Divide the left slope value by the absolute value of the right slope value to obtain the slope ratio of the two sides of the weight.

[0066] It should be noted that the standard deviation of the preset multiple is used to determine the location of the preset unit distance point, and this multiple is set according to the characteristics of the normal distribution.

[0067] A negative slope on the left indicates that the curve is trending downwards to the left of the peak, while a negative slope on the right indicates that the curve is trending downwards to the right of the peak.

[0068] A slope ratio on both sides of the weight distribution that is close to the preset value indicates a symmetrical distribution, while a deviation from the preset value indicates a skewed distribution.

[0069] For the normal distribution curves of impedance in each direction, extract the left slope value, right slope value, and the ratio of the slopes on both sides of each direction channel in the same way.

[0070] The ratio of the first longitudinal channel impedance value to the second longitudinal channel impedance value is used as the longitudinal symmetry index, and the ratio of the transverse channel impedance value to the average of the first and second longitudinal channel impedance values ​​is used as the transverse conduction index.

[0071] It should be noted that a longitudinal symmetry index close to the preset value indicates that the tissues of the left and right feet of the chicken are similar, while a deviation from the preset value indicates that there is asymmetry between the left and right feet.

[0072] The transverse conduction index reflects the efficiency of current conduction between the two feet of a chicken; the lower the value, the greater the transverse conduction resistance.

[0073] The peak weight position, peak weight offset, slope ratio on both sides of the weight, peak impedance position in each direction, peak impedance offset in each direction, slope ratio on both sides of each direction, longitudinal symmetry index, and transverse conduction index are summarized into population state distribution data.

[0074] In one specific embodiment, the analysis of the feeding sound data is carried out as follows: the collected audio signal from the feed tray area is filtered and noise-reduced to remove fan noise and environmental background noise, thereby obtaining the feeding audio signal.

[0075] It should be noted that the filtering and noise reduction process uses a bandpass filter to retain the frequency band of the chicken pecking sound while filtering out low-frequency noise generated by the fan and high-frequency environmental noise.

[0076] The feeding audio signal is processed by frame segmentation. The short-time energy and short-time zero-crossing rate of each frame are extracted. Frames with short-time energy exceeding a preset energy threshold and short-time zero-crossing rate exceeding a preset zero-crossing rate threshold are marked as feeding frames.

[0077] It should be noted that short-time energy is the sum of squares of the sampled audio signal values ​​within a single frame, and short-time zero-crossing rate is the number of times the audio signal waveform crosses the zero axis within a single frame. Pecking sounds are characterized by high short-time energy and high zero-crossing rate. By using dual threshold judgment, pecking sounds can be effectively distinguished from other sounds such as chickens walking or crowing.

[0078] The number of pecking frames in each feed tray per unit time is counted. The number of pecking frames is divided by the preset standard pecking frequency to obtain the frequency score. The average energy value of the pecking frames in each feed tray per unit time is counted. The average energy value is divided by the preset standard energy value to obtain the energy score. The frequency score and the energy score are weighted and summed according to preset weights to obtain the overall flock feeding vitality index.

[0079] It should be noted that the preset standard pecking frequency is the average number of peckings per unit time under normal feeding conditions, and the preset standard energy value is the average energy value of pecking sounds under normal feeding conditions, which is obtained through statistical analysis of pre-collected standard samples.

[0080] The weights of frequency score and energy score are set according to the strength of their correlation with foraging activity.

[0081] In one specific embodiment, the weight distribution status assessment process is as follows: the peak weight position is compared with the standard peak weight position for the current age. When the peak weight position is lower than the standard peak weight position, it is determined that the peak weight has shifted to the left. When the peak weight position is higher than the standard peak weight position, it is determined that the peak weight has shifted to the right.

[0082] It should be noted that the peak standard weight position for the current age is determined based on the standard growth curve of the broiler breed, and different ages correspond to different peak standard positions.

[0083] The ratio of the slopes on both sides of the weight is compared with a preset symmetry range. When the ratio of the slopes on both sides of the weight is within the preset symmetry range, the weight distribution is determined to be symmetrical. When the ratio of the slopes on both sides of the weight exceeds the preset symmetry range, the weight distribution is determined to be skewed.

[0084] It should be noted that the preset symmetry range is determined based on the statistical distribution of the slope ratio of the weight on both sides of the historical normal flock. A leftward shift of the peak weight indicates that the flock is underweight, which may be due to excessive feed restriction or insufficient feed nutrition; a rightward shift of the peak weight indicates that the flock is overweight, which may be due to insufficient feed restriction or excessive feed energy.

[0085] Skewed weight distribution indicates differentiation within the population, with some chickens deviating from the mainstream weight distribution, requiring differentiated feeding based on regional distribution to correct this.

[0086] The results of the peak weight deviation determination and the weight distribution skewness determination are combined to form the weight restriction effect assessment results.

[0087] In one specific embodiment, the impedance distribution state assessment process is as follows: the slope ratio of the impedance on both sides of each directional channel is compared with a preset symmetry range. When the slope ratio of the impedance on both sides of any directional channel exceeds the preset symmetry range, the impedance distribution of that directional channel is determined to be skewed. Combined with the peak offset direction of the impedance of that directional channel, when the peak impedance shifts to the right and the slope value on the right side is less than the slope value on the left side, it is determined that the tissue area corresponding to that directional channel has a dehydration trend. When the peak impedance shifts to the left and the slope value on the left side is less than the slope value on the right side, it is determined that the tissue area corresponding to that directional channel has an edema trend.

[0088] It should be noted that a rightward shift of the impedance peak indicates a general increase in the overall impedance value, which is usually related to tissue dehydration, as reduced water content leads to decreased conductivity and increased impedance. Conversely, a leftward shift of the impedance peak indicates a general decrease in the overall impedance value, which may be related to tissue edema, as increased water content leads to enhanced conductivity and decreased impedance.

[0089] The slope value on the right side is less than that on the left side, indicating that the tail of the right half of the curve is longer, suggesting that some chickens have abnormally high impedance values.

[0090] The longitudinal symmetry index is compared with a preset symmetry index threshold. When the longitudinal symmetry index exceeds the preset symmetry index threshold, it is determined that the tissue state of the left and right feet of the chicken is asymmetrical.

[0091] It should be noted that the longitudinal symmetry index reflects the symmetry of the tissues on the left and right feet. An abnormality in this index may indicate abnormal gait or unilateral leg problems in chickens.

[0092] The transverse conduction index is compared with a preset transverse conduction threshold. When the transverse conduction index is lower than the preset transverse conduction threshold, it is determined that the transverse current conduction in the chicken's body is blocked.

[0093] It should be noted that the transverse conduction index reflects the efficiency of current conduction between the two feet. A decrease in this index may indicate increased tissue resistance in chickens, which is related to malnutrition or metabolic abnormalities.

[0094] The results of the resistance distribution skewness determination, the results of the dehydration trend determination or edema trend determination, the results of the longitudinal symmetry index determination, and the results of the transverse conduction index determination are combined to form the resistance feeding restriction effect assessment results.

[0095] In one specific embodiment, the feeding vitality status assessment process is as follows: the feeding vitality index is compared with a preset lower vitality threshold and a preset upper vitality threshold. When the feeding vitality index is lower than the preset lower vitality threshold, the vitality is determined to be insufficient. When the feeding vitality index is between the preset lower vitality threshold and the preset upper vitality threshold, the vitality is determined to be normal. When the feeding vitality index is higher than the preset upper vitality threshold, the vitality is determined to be vigorous.

[0096] It should be noted that the preset lower limit of vitality is set based on the low quantile of the statistical distribution of feed intake vitality index during historical normal feeding cycles. If it is below this threshold, it indicates that the flock’s feed intake enthusiasm has decreased significantly, which may be due to excessive feed restriction or abnormal flock health.

[0097] The preset upper limit for vitality is set based on the high quantile of the statistical distribution of feed intake vitality index during historical normal feeding cycles. A value exceeding this threshold indicates an abnormally high feed intake activity in the flock, possibly due to insufficient feed restriction leading to excessive hunger. Normal vitality indicates that the stress level of the current feed restriction program on the flock is within an acceptable range.

[0098] Vitality assessment results were used as the assessment results of the impact of vitality restriction on feeding.

[0099] The dynamic feeding correction unit is used to determine whether the current feeding restriction plan needs to be corrected based on the assessment results of the impact of feeding restriction. When correction is needed, the feeding plan is determined based on the population status distribution data and then issued for execution.

[0100] In one specific embodiment, the process of determining the feeding plan is as follows: read the results of the weight restriction impact assessment, the electrical impedance restriction impact assessment, and the vitality restriction impact assessment. When the weight peak in the weight restriction impact assessment shifts to the left, the total feeding amount is increased by a preset step size.

[0101] It should be noted that the preset step length is set according to the feed requirements of the broiler breed and age.

[0102] When increasing the total amount of feed, the increment should be a preset step size to avoid stressing the flock due to excessive adjustments in a single instance.

[0103] When the peak body weight shifts to the right in the assessment results of the impact of feed restriction, the total amount of feed will be reduced by the preset step size.

[0104] When the ratio of the slopes on both sides of the body weight exceeds the preset symmetrical range in the assessment results of the impact of feed restriction, the feed amount is increased in the area where the body weight is lighter when the slope value on the left side is less than the slope value on the right side, and the feed amount is reduced in the area where the body weight is heavier when the slope value on the left side is greater than the slope value on the right side.

[0105] When dehydration trends are observed in any direction of the feed restriction impact assessment results, increase the water supply to the drinking water system; when edema trends are observed in any direction of the feed restriction impact assessment results, decrease the water supply to the drinking water system.

[0106] When the longitudinal symmetry index in the assessment of the impact of feed restriction exceeds the preset symmetry index threshold, electrolyte supplements are added to the feeding program to regulate the electrolyte balance in chickens.

[0107] When the lateral transmission index in the assessment of the impact of feed restriction is lower than the preset lateral transmission threshold, vitamin and trace element supplements are added to the feeding program to improve the tissue health of chickens.

[0108] When the assessment results of insufficient vitality indicate an impact of feed restriction, the total amount of feed should remain unchanged, and the number of feedings per day should be increased in a preset increment.

[0109] The revised total feeding amount, zoned differentiated feeding plan, drinking water adjustment plan, additive supplementation plan, and feeding frequency are combined into a feeding plan and issued for implementation.

[0110] It should be noted that the zoned differentiated feeding scheme is achieved by setting different feeding times at the feed inlets in different areas of the chicken house. The feeding time is extended in areas with lighter body weight and shortened in areas with heavier body weight.

[0111] The water supply of the drinking water system is controlled by adjusting the water pressure or the duration of water supply in the drinking water pipeline.

[0112] Electrolyte supplements, vitamins, and trace element supplements are added through the drinking water system or feed mixing system.

[0113] When feeding frequency increases, the original total daily feed amount is evenly distributed across more meals, and the amount of feed per meal is reduced accordingly.

[0114] according to Figure 2 As shown, the present invention provides a modern feed supply control method for broiler farms, including the following process: Process 1: Multidimensional data acquisition: Collect multidimensional data of the flock and perform standardized processing. The multidimensional data of the flock includes group weight distribution data, individual bioelectrical impedance data and feeding sound data.

[0115] Step 2, Feed Restriction Impact Assessment: Multidimensional normal distribution modeling is performed on the group weight distribution data and individual bioelectrical impedance data to extract the group state distribution data. Feeding sound data is analyzed to obtain the overall group feeding activity index. The group state distribution data is assessed for weight distribution and bioelectrical impedance distribution, and the overall group feeding activity index is assessed for feeding activity status, generating the feed restriction impact assessment results.

[0116] Step 3: Dynamic Feeding Adjustment: Based on the impact assessment results of feed restriction, determine whether the current feed restriction plan needs to be adjusted. If adjustment is required, determine the feeding plan based on the population status distribution data and issue it for implementation.

[0117] The Gaussian function fitting technique, least squares method, bandpass filtering technique, short-time energy calculation and short-time zero-crossing rate calculation described in this invention are all existing technologies and can be found on the Internet, so they will not be described in detail here.

[0118] It should also be noted that the various preset thresholds described in this invention are based on historical feeding data from broiler farms, standard weight curves for broilers of different ages, statistical data on impedance distribution in normal flocks, and a sample database of feeding sounds. These are combined with the feeding and management experience of poultry technicians, and determined by considering the characteristics of the broiler breed, age stage, and stocking density. This approach has a solid scientific basis and practical operability. The preset values ​​specified in the specification, such as preset weight range, preset impedance range, preset multiple, preset energy threshold, preset zero-crossing rate threshold, preset standard pecking frequency, preset standard energy value, preset symmetry range, preset symmetry index threshold, preset lateral conduction threshold, preset lower limit of vitality threshold, preset upper limit of vitality threshold, and preset step size, are all exemplary presets adapted to typical white-feathered broiler farming scenarios. Specific adjustments can be made according to the broiler breed, age stage, and farming scale. The relevant detailed rules are mature existing technologies and will not be elaborated upon here.

[0119] The examples described in this invention are not limited to the specific embodiments listed above. The examples are merely illustrative to facilitate understanding of the invention and do not constitute a limitation on the scope of protection of this invention. Any modifications, equivalent substitutions, etc., made within the spirit and principles of this invention should be included within the scope of protection.

[0120] The above description is merely an example and illustration of the concept of the present invention. Those skilled in the art can make various modifications or additions to the specific embodiments described or use similar methods to replace them, as long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they should all fall within the protection scope of the present invention.

Claims

1. A modern feed supply control system for broiler farms, characterized in that: Includes the following units: The multidimensional data acquisition unit is used to collect multidimensional data of chicken flocks and perform standardized processing. The multidimensional data of chicken flocks includes group weight distribution data, individual bioelectrical impedance data and feeding sound data. The feed restriction impact assessment unit is used to perform multidimensional normal distribution modeling on population weight distribution data and individual bioelectrical impedance data, extract population state distribution data, analyze feeding sound data, obtain the overall feeding activity index, assess the weight distribution status and bioelectrical impedance distribution status of the population state distribution data respectively, assess the feeding activity status of the overall feeding activity index, and generate feed restriction impact assessment results. The dynamic feeding correction unit is used to determine whether the current feeding restriction plan needs to be corrected based on the impact assessment results. When a correction is needed, the feeding plan is determined based on the population status distribution data and then issued for execution.

2. The feed supply control system for modern broiler farms according to claim 1, characterized in that, The multidimensional normal distribution modeling of population weight distribution data and individual bioelectrical impedance data is performed as follows: The group weight distribution data includes the weight values ​​of each chicken, and the individual bioelectrical impedance data includes the multidirectional impedance data of each chicken. The frequency of each chicken's weight value is counted according to the preset weight range. A weight frequency distribution histogram is constructed with the weight value on the horizontal axis and the number of chickens in each weight range on the vertical axis. The weight frequency distribution histogram is fitted with a Gaussian function to obtain the normal distribution curve of the weight. The multi-directional impedance data of each chicken are grouped according to the current injection direction. The impedance value of each directional channel is counted according to the preset impedance interval. The impedance value is plotted on the horizontal axis and the number of chickens in each impedance interval is plotted on the vertical axis to construct the impedance frequency distribution histogram of each directional channel. The histogram of impedance frequency distribution of each directional channel is fitted with a Gaussian function to obtain the normal distribution curve of impedance of each directional channel. The normal distribution curve of body weight and the normal distribution curve of channel impedance in each direction are combined into a multidimensional normal distribution model.

3. The feed supply control system for modern broiler farms according to claim 2, characterized in that, The multi-directional impedance data acquisition process is as follows: Data was collected by a partitioned electrode array laid on the ground of the chicken passageway. The partitioned electrode array is divided into a front zone and a rear zone along the length of the passageway. The front zone is divided into a left front electrode and a right front electrode along the width of the passageway, and the rear zone is divided into a left rear electrode and a right rear electrode along the width of the passageway. When chickens walk through the partitioned electrode array naturally, their left foot steps on the left front electrode and the left rear electrode, and their right foot steps on the right front electrode and the right rear electrode. The partitioned electrode array injects current between each pair of electrodes in a preset sequence and measures the voltage response. The current injection direction includes a first longitudinal channel along the front and back of the chicken's left foot, a second longitudinal channel along the front and back of the chicken's right foot, and a transverse channel between the two feet of the chicken. The first longitudinal channel is composed of the left front electrode and the left rear electrode, the second longitudinal channel is composed of the right front electrode and the right rear electrode, and the transverse channel is composed of the electrode contacted by the left foot and the electrode contacted by the right foot. The impedance values ​​measured by each channel are used as the multi-directional impedance data of the chicken.

4. The feed supply control system for modern broiler farms according to claim 2, characterized in that, The specific extraction process for the population state distribution data is as follows: The mean and standard deviation parameters are extracted from the normal distribution curve of body weight. The mean parameter is used as the position of the peak weight. The current peak weight position is compared with the standard peak weight position of the current age, and the peak weight offset is calculated. The mean and standard deviation parameters of the impedance normal distribution curves of each directional channel are extracted. The mean parameter of each directional channel is used as the impedance peak position of that directional channel. The current impedance peak position of each directional channel is compared with the impedance peak position of the same directional channel in the previous feeding cycle, and the impedance peak offset of each directional channel is calculated. Take a preset unit distance point to the left and right of the peak position of the normal distribution curve of weight. The preset unit distance is the distance of the standard deviation of the peak position offset to both sides by a preset multiple. Calculate the slope of the line connecting the preset unit distance point on the left to the peak position as the left slope value, and calculate the slope of the line connecting the preset unit distance point on the right to the peak position as the right slope value. Divide the left slope value by the absolute value of the right slope value to obtain the slope ratio of the two sides of the weight. For the normal distribution curves of impedance in each direction, extract the left slope value, right slope value, and the ratio of the slopes on both sides of each direction in the same way; The ratio of the first longitudinal channel impedance value to the second longitudinal channel impedance value is used as the longitudinal symmetry index, and the ratio of the transverse channel impedance value to the average of the first and second longitudinal channel impedance values ​​is used as the transverse conduction index. The peak weight position, peak weight offset, slope ratio on both sides of the weight, peak impedance position in each direction, peak impedance offset in each direction, slope ratio on both sides of each direction, longitudinal symmetry index, and transverse conduction index are summarized into population state distribution data.

5. The feed supply control system for modern broiler farms according to claim 4, characterized in that, The analysis of the feeding sound data is as follows: The audio signal from the collected feed tray area is filtered and noise-reduced to remove fan noise and ambient background noise, thus obtaining the feeding audio signal. The feeding audio signal is processed by frame segmentation. The short-time energy and short-time zero-crossing rate of each frame are extracted. Frames with short-time energy exceeding a preset energy threshold and short-time zero-crossing rate exceeding a preset zero-crossing threshold are marked as pecking frames. The number of pecking frames in each feed tray per unit time is counted. The number of pecking frames is divided by the preset standard pecking frequency to obtain the frequency score. The average energy value of the pecking frames in each feed tray per unit time is counted. The average energy value is divided by the preset standard energy value to obtain the energy score. The frequency score and the energy score are weighted and summed according to preset weights to obtain the overall flock feeding vitality index.

6. The feed supply control system for modern broiler farms according to claim 1, characterized in that, The assessment of body weight distribution is conducted as follows: The peak weight position is compared with the standard peak weight position for the current age. If the peak weight position is lower than the standard peak weight position, the peak weight is determined to have shifted to the left. If the peak weight position is higher than the standard peak weight position, the peak weight is determined to have shifted to the right. The ratio of the slopes on both sides of the weight is compared with a preset symmetry range. When the ratio of the slopes on both sides of the weight is within the preset symmetry range, the weight distribution is determined to be symmetrical. When the ratio of the slopes on both sides of the weight exceeds the preset symmetry range, the weight distribution is determined to be skewed. The results of the peak weight deviation determination and the weight distribution skewness determination are combined to form the weight restriction effect assessment results.

7. The feed supply control system for modern broiler farms according to claim 1, characterized in that, The assessment of the electrical impedance distribution is carried out as follows: The slope ratio of the impedance on both sides of each directional channel is compared with a preset symmetry range. When the slope ratio of the impedance on both sides of any directional channel exceeds the preset symmetry range, the impedance distribution of that directional channel is determined to be skewed. Combined with the peak offset direction of the impedance of that directional channel, when the peak impedance shifts to the right and the slope value on the right side is less than the slope value on the left side, it is determined that the tissue area corresponding to that directional channel has a dehydration trend. When the peak impedance shifts to the left and the slope value on the left side is less than the slope value on the right side, it is determined that the tissue area corresponding to that directional channel has an edema trend. The longitudinal symmetry index is compared with a preset symmetry index threshold. When the longitudinal symmetry index exceeds the preset symmetry index threshold, it is determined that the tissue state of the left and right feet of the chicken is asymmetrical. The transverse conduction index is compared with a preset transverse conduction threshold. When the transverse conduction index is lower than the preset transverse conduction threshold, it is determined that the transverse current conduction in the chicken's body is blocked. The results of the resistance distribution skewness determination, the results of the dehydration trend determination or edema trend determination, the results of the longitudinal symmetry index determination, and the results of the transverse conduction index determination are combined to form the resistance feeding restriction effect assessment results.

8. The feed supply control system for modern broiler farms according to claim 1, characterized in that, The specific assessment process for the feeding activity status is as follows: The feeding activity index is compared with the preset lower limit threshold and the preset upper limit threshold. When the feeding activity index is lower than the preset lower limit threshold, the activity is considered insufficient. When the feeding activity index is between the preset lower limit threshold and the preset upper limit threshold, the activity is considered normal. When the feeding activity index is higher than the preset upper limit threshold, the activity is considered vigorous. Vitality assessment results were used as the assessment results of the impact of vitality restriction on feeding.

9. The feed supply control system for modern broiler farms according to claim 1, characterized in that, The specific confirmation process for determining the feeding plan is as follows: Read the results of the weight restriction impact assessment, the electrical impedance restriction impact assessment, and the vitality restriction impact assessment. When the peak weight in the weight restriction impact assessment shifts to the left, adjust the total feeding amount upwards by the preset step size. When the peak body weight shifts to the right in the assessment results of the impact of feed restriction, the total amount of feed will be reduced by the preset step size; When the ratio of the slopes on both sides of the body weight in the assessment of the impact of feed restriction exceeds the preset symmetrical range, the feed amount is increased in the area where the body weight is lighter when the slope value on the left side is less than the slope value on the right side, and the feed amount is reduced in the area where the body weight is heavier when the slope value on the left side is greater than the slope value on the right side. When dehydration trend is observed in any direction of the feed restriction impact assessment results, increase the water supply of the drinking water system; when edema trend is observed in any direction of the feed restriction impact assessment results, decrease the water supply of the drinking water system. When the longitudinal symmetry index in the assessment of the impact of feed restriction exceeds the preset symmetry index threshold, electrolyte supplements are added to the feeding program to regulate the electrolyte balance in chickens. When the lateral transmission index in the assessment of the impact of feed restriction is lower than the preset lateral transmission threshold, vitamin and trace element supplements are added to the feeding program to improve the tissue health of chickens. When the assessment results of insufficient vitality indicate that the feeding restriction has an impact, keep the total amount of feeding unchanged and increase the number of feedings per day in a preset step. The revised total feeding amount, zoned differentiated feeding plan, drinking water adjustment plan, additive supplementation plan, and feeding frequency are combined into a feeding plan and issued for implementation.

10. A feed supply control method utilizing the feed supply control system for modern broiler farms as described in any one of claims 1-9, characterized in that, The process includes the following steps: Process 1: Multidimensional data collection: Collect multidimensional data of the chicken flock and perform standardized processing. The multidimensional data of the chicken flock includes group weight distribution data, individual bioelectrical impedance data and feeding sound data. Step 2, Feed Restriction Impact Assessment: Multidimensional normal distribution modeling is performed on the group weight distribution data and individual bioelectrical impedance data to extract the group state distribution data. Feeding sound data is analyzed to obtain the overall group feeding activity index. The weight distribution state and bioelectrical impedance distribution state are assessed separately for the group state distribution data. The overall feeding activity index is used to assess feeding activity state, and the feed restriction impact assessment results are generated. Step 3: Dynamic Feeding Adjustment: Based on the impact assessment results of feed restriction, determine whether the current feed restriction plan needs to be adjusted. If adjustment is required, determine the feeding plan based on the population status distribution data and issue it for implementation.

Images