A full-process automated black soldier fly large-scale breeding method and system

Abstract

The present application belongs to the technical field of black soldier fly breeding, and discloses a full-process automated black soldier fly large-scale breeding method and system. First, organic solid waste is sequentially treated by impurity removal, crushing, cooking, pulping and sedimentation, the percolate is all returned to the blending link, a compound auxiliary material and a fungicide are added to obtain feeding pulp meeting the breeding standard, black soldier fly eggs are put into breeding boxes, corresponding doses of pulp are fed to the eggs in different growth stages, the eggs are stacked, and then are transported to corresponding breeding subareas; environment data of each subarea is collected to control and maintain the suitable breeding environment, full-process data is collected after each batch of breeding is completed to iterate the biological conversion model, and the control parameters of the next batch are updated, and after the breeding is completed, the worm powder, worm oil and biological organic fertilizer are automatically separated and processed. The method reduces the labor cost of breeding, avoids secondary pollution caused by the discharge of percolate, improves the utilization rate of the site and the stability of the breeding batches, guarantees the uniformity of the growth of the larvae and the quality of the products, and has good economic and environmental benefits.

Description

Technical Field

[0001] This invention belongs to the field of black soldier fly farming technology, and in particular to a fully automated method and system for large-scale black soldier fly farming. Background Technology

[0002] With the annual production of organic solid waste such as kitchen waste, livestock manure, and waste crops in my country exceeding 2 billion tons, the demand for resource-based disposal is urgent. Black soldier fly bioconversion, with its low-carbon and high-value-added technological advantages, has become one of the core pathways for reducing and recycling organic solid waste. At the same time, my country's annual protein shortage for feed has remained above 30 million tons for a long time. High-quality insect protein extracted from black soldier fly bodies can effectively fill the protein supply gap. Large-scale and standardized black soldier fly farming has become the core development direction in the fields of solid waste disposal and agricultural protein supply.

[0003] Currently, the mainstream farming model in the industry is still mainly decentralized family-based artificial farming. This model involves manual labor to complete the entire process of sorting organic solid waste, feeding, environmental management, and harvesting. The equipment investment threshold is low, which is suitable for small-scale free-range farming. However, its labor cost accounts for more than 40% of the total operating cost, and the farming efficiency is low. Core parameters such as temperature, humidity, and feeding amount are all judged by the experience of the farmers, resulting in poor uniformity of black soldier fly individuals and a fluctuation in the infestation rate of more than 20%. In addition, the traditional flat-layer farming model has a site utilization rate of less than 10%, which cannot meet the production capacity requirements of large-scale farming.

[0004] In recent years, semi-automated aquaculture solutions have gradually emerged, but they only optimize a single link. Chinese patent application number 202310652572.7 discloses an insect larvae breeding technology solution based on zoned environmental control. By dividing the breeding stage and controlling environmental parameters in different zones, it solves the technical problem of excessive energy consumption of integrated environmental control. However, it does not cover the need for full-process automation. A large amount of manual labor is still required to complete feeding, transportation, and harvesting operations. It does not set up a mechanism for full disposal of leachate, which is prone to secondary pollution of wastewater. At the same time, it lacks the ability to collect full-process breeding data and iterate models. Core process parameters cannot be continuously optimized. The conversion rate of different batches of materials fluctuates by more than 15%, resulting in poor product quality stability. It also does not cover the automated control of the back-end processing links and cannot achieve standardized operation of the entire process from raw materials to products.

[0005] In summary, existing black soldier fly farming technologies generally suffer from high reliance on manual labor, insufficient full-process control capabilities, lack of data iteration mechanisms, and low standardization. These shortcomings make it difficult to meet the cost reduction, quality improvement, and standardized operation requirements of large-scale farming of tens of thousands of tons or more. There is an urgent need to develop corresponding fully automated farming solutions to address these issues. Summary of the Invention

[0006] To address the shortcomings of existing black soldier fly farming methods, such as high reliance on manual labor, susceptibility to secondary pollution, poor batch stability, and insufficient adaptability to large-scale operations, this invention provides a fully automated method and system for large-scale black soldier fly farming.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: A fully automated method for large-scale black soldier fly farming includes the following steps: S1. Raw material pretreatment: Organic solid waste is sequentially treated by impurity removal, crushing, cooking, pulping and sedimentation. Then, compound auxiliary materials and microbial agents are added and mixed to adjust the moisture content and protein content of the material to the breeding standard to obtain feeding slurry. All leachate separated during the sedimentation process is returned to the material preparation stage to adjust the moisture content of the material and achieve zero leachate discharge. S2. Staged breeding and transportation: Black soldier fly eggs are placed into breeding boxes, and the corresponding amount of feeding slurry is fed according to the growth stage of the black soldier fly. After feeding, the breeding boxes are stacked and transported to the breeding zone corresponding to the current growth stage by automatic transportation equipment. S3. Zonal Environmental Control: Deploy environmental sensors in each aquaculture zone to collect environmental data in real time. Compare the collected data with the preset parameter thresholds of the aquaculture zone to generate control commands to drive the corresponding environmental control equipment to operate, so that each zone maintains an aquaculture environment suitable for the corresponding growth stage. S4. Data Iteration and Optimization: After each batch of black soldier fly farming is completed, the entire process data of that batch of farming is collected, imported into the pre-built biological transformation model to complete the training iteration, and the farming control parameters for the next batch are updated and simultaneously issued for execution. S5. Automated harvesting and processing: After the breeding cycle is completed, the breeding boxes are transferred to the processing area, where fresh black soldier fly larvae and insect sand are separated by screening. The fresh larvae are cleaned, dried, pressed for oil, and packaged to obtain insect powder and insect oil products. The insect sand is decomposed, sterilized, and packaged to obtain bio-organic fertilizer.

[0008] Furthermore, in step S1, the organic solid waste includes any one or more of kitchen waste, poultry and livestock manure, waste crops, Chinese medicine residue, and fruit and vegetable waste; the moisture content of the feed slurry after preparation is 60%~70%, and the protein content is 12%~18%.

[0009] Furthermore, in step S2, the growth stages of the black soldier fly are divided into three consecutive stages: the hatching period, the rapid growth period, and the maturity period; the breeding box is a trayless customized plastic box, which is stacked in multiple layers after feeding.

[0010] Furthermore, in step S3, the parameter thresholds for each aquaculture zone are set differently according to the biological requirements of the corresponding growth stage; the concentrations of ammonia and carbon dioxide in the rapid growth zone are monitored in real time to maintain the concentrations below the safe threshold.

[0011] Furthermore, in step S4, the method for constructing and iterating the biotransformation model is as follows: an initial model is constructed based on historical aquaculture data, with environmental parameters, material parameters, and feeding parameters as input variables, and black soldier fly growth rate, survival rate, and material conversion rate as output variables. The first-generation biotransformation model is generated through neural network training. After each batch of aquaculture is completed, the entire process data of that batch is added to the training sample set, and the model is iteratively optimized. After the iteration is completed, the optimal aquaculture control parameters for the next batch are automatically output and synchronously sent to the control system for execution.

[0012] Furthermore, in step S5, after the breeding boxes are transferred to the processing area, the residual material on the outer wall is first removed by cleaning equipment, and then the stack is disassembled and turned over to pour out the internal material; the material is separated from the fresh insects and insect sand by multi-stage screening equipment; the fresh insects are washed and then inactivated and dehydrated by hot airflow composite drying process; after drying, they are separated into insect powder and insect oil by oil pressing equipment; the insect sand is decomposed and sterilized and then packaged into biological organic fertilizer.

[0013] Furthermore, it includes a pretreatment module, an automated feeding and transfer module, a zoned aquaculture module, an environmental control module, a data iteration module, an automated processing module, and a central control unit; the pretreatment module, the automated feeding and transfer module, the zoned aquaculture module, the environmental control module, the data iteration module, and the automated processing module are all communicatively connected to the central control unit; The pretreatment module is used to treat organic solid waste into feeding slurry that meets aquaculture standards and to achieve full recycling and disposal of leachate; the automated feeding and transfer module is used to complete the egg loading, feeding, stacking, transfer and unstacking operations of the breeding boxes; the zoned breeding module includes multiple physically independent breeding zones, which correspond to the breeding needs of black soldier flies at different growth stages. The environmental control module is used to collect environmental data from each aquaculture zone and drive the environmental control equipment to maintain the stability of the aquaculture environment in each zone; the data iteration module is used to store the aquaculture data throughout the entire process, train the iterative biological transformation model, and synchronize the optimized control parameters to the central control unit; the automated processing module is used to complete the separation, processing, and packaging of fresh insects and insect sand.

[0014] Furthermore, the pretreatment module includes a sorting device, a crusher, a cooking machine, a pulping machine, a sedimentation tank, a mixing tank, and a storage tank connected in sequence; the mixing tank is equipped with an auxiliary material inlet and a microbial agent inlet; the leachate outlet of the sedimentation tank is connected to the liquid inlet of the mixing tank through a return pipeline, which is used to transport the precipitated leachate back to the mixing tank for material blending.

[0015] Furthermore, the zoned aquaculture module includes physically independent incubation zone, rapid growth zone, and maturity zone; the incubation zone is equipped with heating and humidification equipment, the rapid growth zone is equipped with cooling equipment, ventilation equipment, and exhaust gas treatment equipment, and the maturity zone is equipped with dehumidification and ventilation equipment; the environmental control module includes temperature sensors, humidity sensors, ammonia sensors, carbon dioxide sensors, and light sensors distributed in each aquaculture zone, as well as controllers that communicate with each sensor; the controllers receive parameter instructions from the central control unit and drive the corresponding environmental control equipment to operate based on real-time collected data.

[0016] Furthermore, the data iteration module incorporates a data storage unit and a neural network training unit. The data storage unit stores environmental parameters, material parameters, feeding parameters, black soldier fly growth parameters, material conversion rate parameters, and product quality parameters for all batches. The neural network training unit uses the breeding data of newly added batches as training samples to iteratively update the biological transformation model and uploads the optimized control parameters to the central control unit. The automated processing module includes, in sequence, box cleaning equipment, destacking and turning equipment, multi-stage screening equipment, fresh insect washing equipment, hot air drying equipment, oil pressing equipment, packaging equipment, and insect decomposition equipment.

[0017] The present invention has the following beneficial effects: This invention achieves fully automated operation of the entire breeding process without human intervention by linking automated equipment throughout the entire process of raw material pretreatment, feeding and transportation, zoned breeding, and harvesting and processing, and unified scheduling by a central control unit. This effectively solves the defects of existing technologies, such as high dependence on manual labor, high operating costs, and low operating efficiency. At the same time, it adopts a multi-layer stacking breeding mode without pallets, which effectively improves the utilization rate of the unit area and adapts to the needs of large-scale production capacity.

[0018] This invention eliminates the risk of wastewater discharge during the disposal of organic solid waste by designing a full-volume recirculation and disposal of leachate in the raw material pretreatment stage, coupled with a differentiated environmental control system for different zones and stages. At the same time, it precisely matches the biological needs of black soldier flies at each growth stage, effectively solving the defects of existing technologies such as easy generation of secondary pollution, poor environmental control adaptability, resulting in low survival rate and poor growth uniformity of black soldier flies. There is no secondary pollution discharge throughout the process, and the growth stability of black soldier flies is significantly improved.

[0019] This invention utilizes a full-process data collection and biotransformation model iterative optimization mechanism. The full-process data of each batch of aquaculture is automatically imported into the training set to complete model upgrades, dynamically optimizing feeding parameters, environmental control thresholds, and aquaculture cycles. This effectively solves the shortcomings of existing technologies, such as reliance on human experience for core process parameters, large fluctuations in material conversion rates between batches, and unstable product quality. The material conversion rate is continuously optimized, and the uniformity of product quality is fully guaranteed.

[0020] This solution is suitable for the disposal needs of various types of organic solid waste, such as kitchen waste, livestock manure, and agricultural waste. It can be quickly replicated and promoted to various centralized organic solid waste disposal scenarios. It can not only meet the environmental protection requirements of reducing and harmlessly disposing of organic solid waste, but also stably produce insect protein, insect oil, and bio-organic fertilizer products that meet relevant standards. It has both environmental and economic benefits and has extremely high industrial promotion value. Attached Figure Description

[0021] Figure 1 This is a flowchart of a fully automated method for large-scale black soldier fly farming proposed in this invention; Figure 2 This is a schematic diagram of a fully automated large-scale black soldier fly farming system proposed in this invention; Figure 3 This is a line graph showing the number of iterations and prediction accuracy of the biotransformation model proposed in this invention. Figure 4 This is a radar chart showing the environmental regulation adaptation degree of each growth stage proposed in this invention. Detailed Implementation

[0022] 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.

[0023] This invention discloses a fully automated method for large-scale black soldier fly farming, comprising the following steps: S1. Raw material pretreatment: Organic solid waste is first transported to sorting equipment to remove non-degradable impurities such as plastics, metals, and glass. Then, it is crushed by a crusher into pieces with a particle size of no more than 5 cm. It is then sent to a cooking machine and cooked at 121°C for 30 minutes to complete sterilization and maturation. Subsequently, it is prepared into a homogeneous slurry by a pulping machine and transported to a sedimentation tank for static sedimentation for 2 hours to complete the initial separation of solid and liquid.

[0024] The leachate separated during the sedimentation process is entirely transported to the mixing tank via a reflux pipeline, replacing the need for additional water to adjust the material's moisture content. This effectively avoids secondary pollution caused by leachate discharge. Afterward, compound excipients and a special microbial agent are added to the mixing tank and mixed to adjust the material's moisture and protein content to meet black soldier fly farming standards, resulting in a homogeneous feeding slurry. This slurry is then temporarily stored in a 4°C low-temperature storage tank for later use. This entire process is fully sealed and automated, requiring no manual intervention, effectively addressing the high costs associated with manual sorting and preparation in existing technologies.

[0025] S2. Staged breeding and transportation: Black soldier fly eggs of uniform size are quantitatively placed into breeding boxes. Feeding is done with the corresponding amount of feed according to the growth stage of the black soldier fly. After feeding, the breeding boxes are stacked in multiple layers by a stacking device and transported by AGV transport vehicle to the physical independent breeding zone corresponding to the current growth stage. The whole process does not require manual transportation and stacking, effectively solving the defects of low efficiency and insufficient site utilization of existing technologies.

[0026] S3. Zonal Environmental Control: Multiple types of environmental sensors are distributed and deployed in each breeding zone to collect environmental data in real time. The collected data is compared with the preset parameter thresholds of the breeding zone in real time to generate control commands to drive the corresponding environmental control equipment to operate. This ensures that each zone maintains the optimal breeding environment for the corresponding growth stage, eliminating the need for manual adjustment of environmental parameters based on experience. This effectively solves the shortcomings of existing technologies, such as poor adaptability of environmental control and low uniformity of larval growth.

[0027] S4. Data Iteration and Optimization: After each batch of black soldier fly farming is completed, the system automatically collects material, environmental, feeding, and growth data for the entire farming process of that batch. The data is then imported into a pre-built biological transformation model to complete training and iteration. The farming control parameters for the next batch are updated and simultaneously sent to the central control unit for execution. This eliminates the need for manual summarization of farming experience to adjust parameters, effectively solving the shortcomings of existing technologies such as large fluctuations in farming effects between batches and unstable product quality.

[0028] S5. Automated Harvesting and Processing: After the breeding cycle ends, the breeding boxes are automatically transferred to the processing area by AGV transport vehicles. The black soldier fly larvae and insect excrement are separated by automated screening equipment. The fresh larvae are automatically cleaned, dried, pressed for oil, and packaged to obtain insect powder and insect oil products that meet feed standards. The insect excrement is decomposed, sterilized, and packaged to obtain bio-organic fertilizer. The entire processing process is automated, and the core processes do not require manual operation, effectively solving the defects of high labor costs and low efficiency in the existing harvesting and processing technology.

[0029] This invention also discloses the selectable range of raw materials and the parameter thresholds of the feeding slurry. The organic solid waste includes any one or more of kitchen waste, poultry and livestock manure, waste crops, Chinese medicine residue, and fruit and vegetable waste, which can be adapted to the organic solid waste disposal needs of different regions. The moisture content of the prepared feeding slurry is 60%~70%, and the protein content is 12%~18%. This parameter range matches the feeding preferences of black soldier fly larvae, avoiding the problem of poor larval growth caused by unsuitable material parameters, and effectively solving the defect of poor material adaptability in the prior art.

[0030] This invention also discloses the division rules for the growth stages of black soldier flies and the selection of breeding boxes. The growth stages of black soldier flies are divided into three consecutive stages: the hatching period, the rapid growth period, and the maturity period. The boundaries of each stage are clear and the differences in biological requirements are obvious, which facilitates zoned management. The breeding boxes are made of palletless customized food-grade PP plastic boxes. Each box can bear a weight of no less than 50 kg. After feeding, they can be stacked up to 12 layers high. The utilization rate of the unit area is effectively improved compared with flat-layer breeding, which effectively solves the defects of low site utilization and poor adaptability to large-scale production in existing technologies.

[0031] This invention also discloses parameter setting rules for breeding zones. The parameter thresholds of each breeding zone are set differently according to the biological needs of the corresponding growth stage, avoiding energy waste caused by uniform control. The ammonia and carbon dioxide concentrations in the rapid growth zone are monitored in real time and kept below the safe threshold to avoid the accumulation of harmful gases affecting larval growth. This effectively solves the defects of existing technologies such as insufficient environmental control and low larval survival rate.

[0032] This invention also discloses a method for constructing and iterating a biotransformation model. An initial model is constructed based on historical aquaculture data, with environmental parameters, material parameters, and feeding parameters as input variables, and black soldier fly growth rate, survival rate, and material conversion rate as output variables. The first-generation biotransformation model is generated through neural network training. After each batch of aquaculture is completed, the entire process data of that batch is added to the training sample set to iteratively optimize the model. After the iteration is completed, the optimal aquaculture control parameters for the next batch are automatically output and simultaneously sent to the control system for execution. The entire model iteration process does not require manual intervention, and the efficiency of control parameter optimization is far higher than that of manual experience summarization, effectively solving the defect of poor batch stability in existing technologies.

[0033] This invention also discloses a specific process for automated harvesting and processing. After the breeding boxes are transported to the processing area, they are first cleaned by a high-pressure spray cleaning device to remove residual materials from the outer wall, and then the internal materials are poured out by a destacking device layer by layer. The materials are separated into fresh insects and insect excrement by a three-stage vibrating screening device. The fresh insects are cleaned by bubbles and then inactivated and dehydrated by a 120°C hot airflow composite drying process. After drying, they are separated into insect powder and insect oil by a cold pressing oil extraction device. The insect excrement is treated with aerobic composting and sterilization and then packaged into bio-organic fertilizer. The entire process is automated, and the product quality is stable and controllable, effectively solving the defects of low harvesting and processing efficiency and unstable product quality in existing technologies.

[0034] This invention also discloses a fully automated large-scale black soldier fly farming system for executing the aforementioned fully automated large-scale black soldier fly farming method. The system includes a pretreatment module, an automated feeding and transfer module, a zoned farming module, an environmental control module, a data iteration module, an automated processing module, and a central control unit. The pretreatment module, automated feeding and transfer module, zoned farming module, environmental control module, data iteration module, and automated processing module are all communicatively connected to the central control unit. The pretreatment module processes organic solid waste into feed slurry that meets aquaculture standards and enables full recycling and disposal of leachate. The automated feeding and transfer module handles egg loading, feeding, stacking, transfer, and unstacking operations in the aquaculture boxes. The zoned aquaculture module includes multiple physically independent aquaculture zones, each corresponding to the aquaculture needs of black soldier flies at different growth stages. The environmental control module collects environmental data from each aquaculture zone and drives environmental control equipment to maintain a stable aquaculture environment in each zone. The data iteration module stores aquaculture data throughout the entire process, trains and iterates the biotransformation model, and synchronizes the optimized control parameters to the central control unit. The automated processing module separates, processes, and packages fresh insects from insect waste. The entire system is centrally controlled and managed, eliminating the need for manual intervention in the operation of each module and effectively addressing the high reliance on manual labor in existing technologies.

[0035] This invention also discloses the specific structure of the pretreatment module, which includes a sorting device, a crusher, a cooking machine, a pulping machine, a sedimentation tank, a mixing tank, and a storage tank connected in sequence. The mixing tank is equipped with an auxiliary material inlet and a microbial agent inlet. The leachate outlet of the sedimentation tank is connected to the liquid inlet of the mixing tank through a corrosion-resistant return pipeline, which is used to transport the precipitated leachate back to the mixing tank for material blending. There is no wastewater discharge throughout the process, which effectively solves the defect of the prior art that is prone to secondary pollution.

[0036] This invention also discloses the specific structures of the zoned aquaculture module and the environmental control module. The zoned aquaculture module includes three physically independent zones: an incubation zone, a rapid growth zone, and a maturity zone. The incubation zone is equipped with heating and humidification equipment, the rapid growth zone is equipped with cooling, ventilation, and exhaust gas treatment equipment, and the maturity zone is equipped with dehumidification and ventilation equipment. The environmental control module includes temperature sensors, humidity sensors, ammonia sensors, carbon dioxide sensors, and light sensors distributed in each aquaculture zone, as well as controllers that communicate with each sensor. The controllers receive parameter instructions from the central control unit and drive the corresponding environmental control equipment to operate based on real-time collected data. Each zone is independently controlled, resulting in lower energy consumption and higher control accuracy, effectively solving the shortcomings of poor environmental control adaptability in existing technologies.

[0037] This invention also discloses the specific structures of the data iteration module and the automated processing module. The data iteration module has a built-in data storage unit and a neural network training unit. The data storage unit is used to store environmental parameters, material parameters, feeding parameters, black soldier fly growth parameters, material conversion rate parameters, and product quality parameters for all batches. The neural network training unit uses the breeding data of newly added batches as training samples to complete the iterative update of the biological transformation model and uploads the optimized control parameters to the central control unit. The automated processing module includes a box cleaning device, a stacking and turning device, a multi-stage screening device, a fresh insect washing device, a hot air drying device, an oil pressing device, a packaging device, and an insect and sand composting device connected in sequence. The entire module operates automatically, which greatly improves the processing efficiency and product stability, effectively solving the defects of low harvesting and processing efficiency and unstable product quality in the existing technology.

[0038] Specific implementation plan for core parameters 1. Specific formula for compound excipients The compound excipient consists of 30 parts by weight of wheat bran, 25 parts by weight of rice bran, 40 parts by weight of corn stalk powder, and 5 parts by weight of shell powder. All components are pulverized to 80 mesh and then mixed in a horizontal ribbon mixer for 15 minutes to obtain a homogeneous excipient. The addition ratio is 8% to 12% of the wet weight of organic solid waste, and it is used to adjust the carbon-nitrogen ratio and air permeability of the material.

[0039] 2. Specific plans for specialized microbial agents The specialized microbial agent is a compound of Bacillus subtilis, Saccharomyces cerevisiae, and Aspergillus niger in a colony count ratio of 3:2:1. All strains are sourced from the China General Microbiological Culture Collection Center. Activation is performed using LB liquid medium and cultured for 24 hours at 37°C and 180 rpm on a shaker. The resulting bacterial solution has an effective viable count of no less than 10⁻⁶. 9 The concentration of CFU / mL is added at a ratio of 0.3% to 0.5% of the total weight of the material to accelerate material decomposition and improve the feeding efficiency of larvae.

[0040] 3. Specific Scheme for Biotransformation Model The biotransformation model employs a four-layer backpropagation (BP) neural network, consisting of one input layer, two hidden layers, and one output layer. The input layer has 12 neurons, corresponding to the following input variables: ambient temperature, ambient humidity, light intensity, ammonia concentration, carbon dioxide concentration, material moisture content, material protein content, microbial agent dosage, daily feed amount, feeding frequency, egg release amount, and number of culture days. The hidden layer consists of two layers, each with 24 neurons, and the activation function is ReLU. The output layer consists of three neurons, with the corresponding output variables being the daily growth rate of black soldier fly, larval survival rate, and material conversion rate. The training algorithm uses the Adam optimization algorithm, with an initial learning rate of 0.001, a maximum number of iterations of 1000, and an initial training sample size of 1200 batches of historical aquaculture data.

[0041] 4. Specific ranges of environmental parameters at each stage During the incubation period, the temperature should be controlled at 28~32℃, the humidity at 75%~85%, and the light intensity at 0~10 lux; during the rapid growth period, the temperature should be controlled at 25~30℃, the humidity at 65%~75%, the ammonia concentration at no more than 20 ppm, and the carbon dioxide concentration at no more than 1500 ppm; during the maturity period, the temperature should be controlled at 22~28℃, the humidity at 50%~60%, and the light intensity at 100~200 lux.

[0042] Device connection and overall workflow sequence The material flow of the pretreatment module is as follows: organic solid waste enters the sorting equipment, crusher, cooking machine, pulper, sedimentation tank, mixing tank and storage tank in sequence. The leachate outlet at the bottom of the sedimentation tank is connected to the liquid inlet of the mixing tank through a corrosion-resistant centrifugal pump and a flow solenoid valve. The leachate return flow rate is automatically adjusted by the central control unit according to the material moisture content requirements.

[0043] The automated feeding and transfer module operates as follows: empty breeding boxes are fed → black soldier fly eggs are added in a quantitative manner → initial feeding → stacking (8 boxes per layer, stacked to a height of 12 layers) → AGV transport vehicle transfers to the corresponding breeding zone → after the breeding cycle ends, AGV transport vehicle transfers to the processing area → stacking is carried out layer by layer. All actions are scheduled uniformly by the central control unit.

[0044] The automated processing module connection is as follows: the breeding boxes enter the box cleaning equipment → the destacking and turning equipment → the three-stage vibrating screening equipment → the fresh insects enter the washing, drying, oil pressing and packaging equipment, and the insect sand enters the composting and sterilization equipment and the packaging equipment. The entire process of material transportation is automatic and does not require manual transfer.

[0045] Scenario Examples Scenario 1: Application of County-level Organic Solid Waste Treatment Center Application scenario description: A county-level organic solid waste treatment center generates 7 tons of kitchen waste and 3 tons of fruit and vegetable waste daily. The existing technology uses artificial black soldier fly breeding, which has problems such as high labor costs, easy leakage of leachate polluting the surrounding environment, and large fluctuations in batch infestation rates.

[0046] Technical adaptation details in the scenario: The pretreatment module is adapted to the characteristics of kitchen waste with high moisture content, the leachate return ratio is set to 20%~30%, the amount of compound auxiliary materials added is set to 8%, the amount of special microbial agent added is set to 0.3%, the biological transformation model is preloaded with a training subset dedicated to kitchen waste breeding, the AGV transport vehicle transfer route is planned in advance according to the site layout, and the feeding interval is set to 24 hours.

[0047] The specific manifestations of the defect resolution in the scenario: Only one maintenance personnel is needed to conduct inspections twice a day throughout the entire process. No manual intervention is required in the breeding operation, and the labor cost is reduced by about 62% compared with the original model. The leachate is completely recycled and disposed of without external discharge. The material conversion rate of 12 consecutive batches of breeding does not fluctuate by more than 3%, which is far lower than the 22% of the original model. This effectively solves the defects of the original model, such as high dependence on manual labor, easy generation of secondary pollution, and poor batch stability.

[0048] System operation process: After the raw materials are weighed upon entering the factory, they automatically enter the pretreatment module. The central control unit automatically adjusts the amount of auxiliary materials and microbial agents added and the leachate return ratio according to the raw material detection data. The prepared feeding slurry is automatically transported to the feeding station for feeding. The breeding boxes are transferred to the corresponding zones by AGV transport vehicles. Environmental sensors collect data in real time to regulate the zone environment. After each batch of breeding is completed, the data is automatically uploaded to the data iteration module to complete the model update. After the breeding cycle ends, the boxes are automatically transferred to the processing area to complete product processing.

[0049] Scenario 2: Application in supporting large-scale dairy farms Application scenario description: A large-scale dairy farm with 1,000 cows produces 50 tons of cow manure per day. The existing technology adopts the composting disposal mode, which has the problems of high manure disposal cost, long composting cycle and inability to generate high-value benefits.

[0050] Technical adaptation details in the scenario: The pretreatment module adds a coarse fiber crushing unit to crush the coarse fiber in cow manure to a particle size of no more than 1 cm. The amount of compound auxiliary materials added is set to 12%, the carbon source of the supplementary materials is added, the amount of special microbial agent added is set to 0.5%, the ventilation frequency of the rapid growth period is set to 2 times per hour, the biological transformation model is preloaded with a cow manure breeding-specific training subset, and the output insect meal quality parameters are synchronously connected to the feed processing system of the farm.

[0051] The specific manifestations of the solution to the defects in the scenario: cow manure is completely disposed of on-site without the need for external transportation, and the disposal cost is reduced by about 48% compared with the composting model. The breeding cycle is only 15 days, which is much shorter than the 60 days of composting. The protein content of the produced insect meal is stable at over 60%, which can effectively replace soybean meal for feed. It is estimated that it can save the farm more than 1.2 million yuan in feed costs annually, effectively solving the defects of low disposal efficiency, poor returns, and insufficient adaptability to large scale of the original model.

[0052] System operation process: After the cow manure is automatically collected, it is transported to the pretreatment module. The central control unit automatically adjusts the amount of auxiliary materials added and the material parameters. After the feed slurry is prepared, it is automatically fed. The AGV transport vehicle transfers the breeding boxes according to the preset route. The environmental control module automatically maintains the stability of the zoned environment. Each batch of insect powder is directly transported to the feed processing workshop. The insect sand is directly returned to the field as organic fertilizer. The whole process does not require manual intervention.

[0053] Core Algorithm Formulas and Explanations The loss function used for training the biological transformation model is: Where Loss is the model training loss value, and n is the batch size of the training samples. This is the weighted value of the actual output of the i-th breeding batch (calculated by weighting the actual growth rate, survival rate, and material conversion rate according to preset weights, for example, the weights of the three can be set to 0.3, 0.4, and 0.3 respectively). This is the weighted value for the model prediction output of the i-th batch of aquaculture. This formula is used to calculate the deviation between the model prediction result and the actual aquaculture result, driving the iterative update of model parameters, continuously narrowing the gap between the aquaculture control parameters and the optimal value, and effectively solving the defect of large fluctuations in batch aquaculture results caused by the reliance on manual experience to adjust parameters in existing technologies.

[0054] Complete Aquaculture Example This embodiment has a processing capacity of 20 tons of mixed organic solid waste per day, of which kitchen waste accounts for 60%, fruit and vegetable waste accounts for 30%, and traditional Chinese medicine residue accounts for 10%. The amount of compound auxiliary materials added is 10% of the wet weight of organic solid waste, and the amount of special microbial agent added is 0.4% of the total weight of materials. The moisture content of the feed slurry after preparation is 65% and the protein content is 15%. 0.5 grams of black soldier fly eggs are added to each box. The breeding boxes are stacked 12 layers high. The incubation period is 3 days, the rapid growth period is 10 days, and the maturity period is 2 days. The total breeding cycle is 15 days. Each batch can produce 2.8 tons of fresh black soldier fly larvae and 7.2 tons of fly larvae sand. The material conversion rate is 14%. The protein content of the fly larvae powder is 62%, the crude fat content of the fly larvae oil is 80%, and the organic matter content of the fly larvae sand is 75%. All indicators meet the relevant national standards.

[0055] refer to Figure 3This figure illustrates the iterative optimization effect of the biotransformation model as the number of farming batches accumulates. The initial model was trained based on 1200 batches of historical data, with an initial accuracy of 72%. As new farming data was continuously added to the training set, the accuracy reached 95% after 800 iterations, and subsequent growth tended to stabilize. The continuous improvement in the model's prediction accuracy can continuously narrow the gap between farming control parameters and optimal values, effectively solving the shortcomings of existing technologies that rely on manual experience to adjust parameters, resulting in large fluctuations in batch farming effects and unstable product quality. This provides reliable algorithmic support for the continuous optimization of farming parameters.

[0056] refer to Figure 4 This figure shows the environmental control adaptability of the three breeding zones. The adaptability to the core needs of each stage is at a high level. Among them, the adaptability of light and humidity during the hatching period is higher, matching the low light and high humidity requirements of egg hatching; the adaptability of harmful gas control during the rapid growth period is the highest, meeting the harmful gas control requirements during the vigorous feeding stage of larvae; and the adaptability of temperature and light during the maturity period is higher, matching the environmental needs of the prepupal stage of larvae. This verifies the scientific nature of zoned differentiated control and effectively solves the defects of poor adaptability of uniform environmental control and low uniformity of larval growth in existing technologies, providing a reference for further optimization of environmental control parameters.

[0057] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A fully automated method for large-scale black soldier fly farming, characterized in that, Includes the following steps: S1. Raw material pretreatment: Organic solid waste is treated sequentially by removing impurities, crushing, cooking, pulping and sedimentation. Compound auxiliary materials and microbial agents are added and mixed to adjust the moisture content and protein content of the material to the breeding standard to obtain feeding slurry. S2. Staged breeding and transportation: Black soldier fly eggs are placed into breeding boxes, and the corresponding amount of feeding slurry is fed according to the growth stage of the black soldier fly. After feeding, the breeding boxes are stacked and transported to the breeding zone corresponding to the current growth stage by automatic transportation equipment. S3. Zonal Environmental Control: Deploy environmental sensors in each aquaculture zone to collect environmental data in real time. Compare the collected data with the preset parameter thresholds of the aquaculture zone to generate control commands to drive the corresponding environmental control equipment to operate. S4. Data Iteration and Optimization: After each batch of black soldier fly farming is completed, the entire process data of that batch of farming is collected, imported into the pre-built biological transformation model to complete the training iteration, and the farming control parameters for the next batch are updated and simultaneously issued for execution. S5. Automated harvesting and processing: After the breeding cycle is completed, the breeding boxes are transferred to the processing area, where fresh black soldier fly larvae and insect sand are separated by screening. The fresh larvae are cleaned, dried, pressed for oil, and packaged to obtain insect powder and insect oil products. The insect sand is decomposed, sterilized, and packaged to obtain bio-organic fertilizer.

2. The fully automated large-scale black soldier fly farming method according to claim 1, characterized in that, In step S1, the organic solid waste includes any one or more of the following: kitchen waste, poultry and livestock manure, waste crops, Chinese medicine residue, and fruit and vegetable waste; the moisture content of the feed slurry after preparation is 60%~70%, and the protein content is 12%~18%.

3. The fully automated large-scale black soldier fly farming method according to claim 1, characterized in that, In step S2, the growth stages of black soldier flies are divided into three consecutive stages: the hatching period, the rapid growth period, and the maturity period. The breeding box is a trayless custom plastic box, which is stacked in multiple layers after feeding.

4. The fully automated large-scale black soldier fly farming method according to claim 3, characterized in that, In step S3, the parameter thresholds for each aquaculture zone are set differently according to the biological requirements of the corresponding growth stage; the concentrations of ammonia and carbon dioxide in the rapid growth zone are monitored in real time to maintain the concentrations below the safe threshold.

5. The fully automated large-scale black soldier fly farming method according to claim 1, characterized in that, In step S4, the construction and iteration method of the biotransformation model is as follows: an initial model is constructed based on historical aquaculture data, using environmental parameters, material parameters, and feeding parameters as input variables, and black soldier fly growth rate, survival rate, and material conversion rate as output variables. The first-generation biotransformation model is generated through neural network training. After each batch of aquaculture is completed, the entire process data of that batch is added to the training sample set, and the model is iteratively optimized. After the iteration is completed, the optimal aquaculture control parameters for the next batch are automatically output and synchronously sent to the control system for execution.

6. The fully automated large-scale black soldier fly farming method according to claim 1, characterized in that, In step S5, after the breeding boxes are transferred to the processing area, the residual material on the outer wall is removed by the cleaning equipment, and then the internal material is poured out by dismantling and turning over. The material is separated into fresh insects and insect sand by multi-stage screening equipment. After the fresh insects are washed, they are inactivated and dehydrated by hot airflow composite drying process. After drying, they are separated into insect powder and insect oil by oil pressing equipment. The insect sand is packaged into biological organic fertilizer after being decomposed and sterilized.

7. A fully automated large-scale black soldier fly farming system, used to execute the fully automated large-scale black soldier fly farming method according to any one of claims 1 to 6, characterized in that, It includes a pretreatment module, an automated feeding and transfer module, a zoned aquaculture module, an environmental control module, a data iteration module, an automated processing module, and a central control unit; the pretreatment module, the automated feeding and transfer module, the zoned aquaculture module, the environmental control module, the data iteration module, and the automated processing module are all communicatively connected to the central control unit.

8. The fully automated large-scale black soldier fly farming system according to claim 7, characterized in that, The pretreatment module includes a sorting device, a crusher, a cooking machine, a pulping machine, a sedimentation tank, a mixing tank, and a storage tank connected in sequence. The mixing tank is equipped with an auxiliary material inlet and a microbial agent inlet. The leachate outlet of the sedimentation tank is connected to the liquid inlet of the mixing tank through a return pipeline, which is used to transport the precipitated leachate back to the mixing tank for material blending.

9. The fully automated large-scale black soldier fly farming system according to claim 7, characterized in that, The zoned aquaculture module includes three physically independent zones: an incubation zone, a rapid growth zone, and a maturity zone. The incubation zone is equipped with heating and humidification equipment, the rapid growth zone is equipped with cooling, ventilation, and exhaust gas treatment equipment, and the maturity zone is equipped with dehumidification and ventilation equipment. The environmental control module includes temperature sensors, humidity sensors, ammonia sensors, carbon dioxide sensors, and light sensors distributed in each aquaculture zone, as well as controllers that communicate with each sensor. The controllers receive parameter commands from the central control unit and drive the corresponding environmental control equipment to operate based on real-time collected data.

10. The fully automated large-scale black soldier fly farming system according to claim 7, characterized in that, The data iteration module has a built-in data storage unit and a neural network training unit. The data storage unit is used to store environmental parameters, material parameters, feeding parameters, black soldier fly growth parameters, material conversion rate parameters, and product quality parameters for all batches. The neural network training unit uses the breeding data of newly added batches as training samples to complete the iterative update of the biological transformation model and uploads the optimized control parameters to the central control unit. The automated processing module includes, in sequence, box cleaning equipment, destacking and turning equipment, multi-stage screening equipment, fresh insect washing equipment, hot air drying equipment, oil pressing equipment, packaging equipment, and insect sand composting equipment.

Images