To further improve the simulation and estimation accuracy of the supply and demand process of agricultural products, a large number of agricultural data at the national and provincial levels since 1980 were used as the basic research sample, including production, planted area, food consumption, industrial consumption, feed consumption, seed consumption, import, export, price, GDP, population, urban population, rural population, weather and so on, by fully considering the impact factors of agricultural products such as varieties, time, income and economic development, a multi-agricultural products supply and demand forecasting model based on long short-term memory neural network (LSTM) was constructed in this study. The general thought of supply and demand forecasting model is packaging deep neural network training model as an I/O-opening modular model, reserving control interface for input of outside data, and realizing the indicators forecasting of supply and demand and matrixing of balance sheet. The input of model included forecasting balance sheet data of agricultural products, annual price data, general economic data, and international currency data since 2000. The output of model was balance sheet data of next decade since forecasting time. Under the premise of fully considering the mechanical constraints, the model used the advantages of deep learning algorithms in nonlinear model analysis and prediction to analyze and predict supply and demand of 9 main types of agricultural products, including rice, wheat, corn, soybean, pork, poultry, beef, mutton, and aquatic products. The production forecast results of 2019－2021 based on this model were compared and verified with the data published by the National Bureau of Statistics, and the mean absolute percentage error was 3.02%, which meant the average forecast accuracy rate of 2019－2021 was 96.98%. The average forecast accuracy rate was 96.10% in 2019, 98.26% in 2020, and 96.58% in 2021, which shows that with the increase of sample size, the prediction effect of intelligent learning model would gradually get better. The forecasting results indicate that the multi-agricultural supply and demand prediction model based on LSTM constructed in this study can effectively reflect the impact of changes in hidden indicators on the prediction results, avoiding the uncontrollable error introduced by manual experience intervention. The model can provide data production and technical support such as market warning, policy evaluation, resource management and public opinion analysis for agricultural production and management and macroeconomic regulation, and can provide intelligent technical support for multi-regional and inter-temporal agricultural outlook work by monitoring agricultural operation data in a timely manner.
Artificial intelligence (AI) assisted planting can improve in the precise management of protected horticultural crops while also alleviating the increasingly prevalent problem of labor shortage. As a typical representative of labor-intensive industries, the strawberry industry has a growing need for intelligent technology. To assess the regulatory effects of various AI strategies and key technologies on strawberry production in greenhouse, as well as provide valuable references for the innovation and industrial application of AI in horticultural crops, four AI planting strategies were evaluated. Four 96 m2 modern greenhouses were used for planting strawberry plants. Each greenhouse was equipped with standard sensors and actuators, and growers used artificial intelligence algorithms to remotely control the greenhouse climate and crop growth. The regulatory effects of four different AI planting strategies on strawberry growth, fruit yield and qualitywere compared and analyzed. And human-operated cultivation was taken as a reference to analyze the characteristics, existing problems and shortages. Each AI planting strategy simulated and forecast the greenhouse environment and crop growth by constructing models. AI-1 implemented greenhouse management decisions primarily through the knowledge graph method, whereas AI-2 transferred the intelligent planting model of Dutch greenhouse tomato planting to strawberry planting. AI-3 and AI-4 created growth and development models for strawberries based on World Food Studies (WOFOST) and Product of Thermal Effectiveness and Photosynthesis Active Radiation (TEP), respectively. The results showed that all AI supported strategy outperformed a human-operated greenhouse that served as reference. In comparison to the human-operated cultivation group, the average yield and output value of the AI planting strategy group increased 1.66 and 1.82 times, respectively, while the highest Return on Investment increased 1.27 times. AI can effectively improve the accuracy of strawberry planting management and regulation, reduce water, fertilizer, labor input, and obtain higher returns under greenhouse production conditions equipped with relatively complete intelligent equipment and control components, all with the goal of high yield and quality. Key technologies such as knowledge graphs, deep learning, visual recognition, crop models, and crop growth simulators all played a unique role in strawberry AI planting. The average yield and Return on Investment (ROI) of the AI groups were greater than those of the human-operated cultivation group. More specifically, the regulation of AI-1 on crop development and production was relatively stable, integrating expert experience, crop data, and environmental data with knowledge graphs to create a standardized strawberry planting knowledge structure as well as intelligent planting decision-making approach. In this study, AI-1 achieved the highest yield, the heaviest average fruit weight, and the highest ROI. This group's AI-assisted strategy optimized the regulatory effect of growth, development, and yield formation of strawberry crops in consideration of high yield and quality. However, there are still issues to be resolved, such as the difficulty of simulating the disturbance caused by manual management and collecting crop ontology data.
In order to get blueberries goes on sale in advance and obtain greater economic benefits, southern blueberries were moved to an intelligent greenhouse with controllable environment for experimental production. The early maturing production control technology of southern blueberry intelligent greenhouse was explored and studied. First, a detailed and comprehensive investigation and summary were conducted on the production factors of blueberry soilless cultivation, such as the production characteristics of various blueberry varieties, the pH and composition of the substrate, the key points of water and fertilizer irrigation, and the scope of the microenvironment climate. Then, the existing Venlo-type greenhouse was deployed for blueberry production, and the geography, climate and internal structural conditions of the greenhouse were briefly described, and the greenhouse blueberry full-cycle control goal was planned. Finally, the production control system was designed and implemented based on the Internet of Things technology, and the overall framework of the software layer, the hardware layer and the cloud were introduced. Based on multi-factor coordinated control model of greenhouse environment, according to the characteristics of blueberry growth environment, a set of blueberry greenhouse multi-factor coordinated control algorithms were proposed and used for environmental regulation. The experimental greenhouse is located in the southeast of Huaqiao Town, Kunshan city, Suzhou city, Jiangsu province. It has been verified that the overall control system has a significant effect, and the first wave of fruits was harvested in early May 2021, making the southern variety of blueberry enter the fruit picking period nearly one month earlier. Compared with the blueberry plants without cold storage, the yields per plant of "Star" "Emerald" "Lanmei No. 1", and "Coast" after cold storage increased by 51.5%, 85.5%, 43.8%, and 94.7%, respectively, and the weight of each fruit was increased 10.9%, 7.2%, 2.6%, and 5.3%, respectively. Experiments proved that the use of multi-factor coordinated control algorithms for regulation can increase the yield and quality of blueberries and achieve significant economic benefits and provide a demonstration for the industrialization of blueberry plants in southern greenhouses to promote early maturity production and management.
Irrigation is an important factor that affects crop yield. In order to control irrigation of facility crops more effectively and accurately, this study took "Zhangji" strawberry as an example, introduced crop real-time growth characteristics into irrigation decision-making, and combined Penman-Monteith (P-M) model and knowledge reasoning to study the irrigation of strawberry. In the first step, the influencing factors and expert experience in identifying strawberry growth period of "Zhangji" strawberry irrigation were standardized, and the strawberry irrigation data structure based on Resource Description Framework (RDF) was established. The second step was to collect expert experience of strawberry irrigation according to the standardized knowledge structure model. Firstly, all data were unified into structured data, and then were stored in *.csv format together with expert experience, and strawberry irrigation knowledge map based on Neo4j was constructed. The third step was to collect the environmental data and plant data of strawberry in each growth period. The fourth step was using P-M model to calculate the initial irrigation value of strawberry, and then adjusted the initial irrigation value by knowledge reasoning.The fifth step was to conduct experimental planting and evaluate the sampled fruits. In knowledge reasoning, irrigation adjustment strategies of each expert was different. In strawberry irrigation experiment based on P-M model and path sorting algorithm, a group of irrigation reasoning values with the highest probability value were selected to adjust irrigation with the goal of maximizing strawberry yield. The experimental results showed that under the condition of harvesting at a specified time, The total fruit yield, average fruit yield per plant and average fruit weight percentage increased by 2478.5 g, 20.65 g and 12.15% (average fruit weight increased by 1.65 g per fruit) based on P-M model and path sorting algorithm compared with traditional P-M model, respectively. First, on the basis of P-M model, the yield-first irrigation adjustment strategy was adopted. Based on knowledge reasoning, the irrigation frequency and amount were adjusted timely according to the crop growth situation, which improved the yield. Second, under the condition of harvesting and recording yield at a specified time, the experiment accurately controlled the growth period to ensure early fruit ripening, while the other three groups of fruits were not fully mature and the yield of immature fruits were not calculated. Under the method of strawberry irrigation based on Penman-Monteith model and path sorting algorithm, the fruit was picked within a fixed time and reached 0.39 kg/cm2, which increased by 0.1 kg/cm2. Because the planting goal of this study was yield first, only the influence of irrigation on yield was considered. The experimental resulted show that the irrigation method based on model and knowledge reasoning could improve the yield of strawberry, and can provide a new idea for precise irrigation.
The digital elevation model (DEM) of maize population in field was constructed by using optical imaging equipment mounted on unmanned aerial vehicle (UAV) to study the accuracy difference of maize population height monitoring at different growth stages. Three cultivars and eight sowing date treatments were set up to structure maize population with different plant heights. A multi-rotor UAV with high-definition digital camera and multispectral imaging sensor was used to take RGB images and multispectral images in the experiment area on July 25th and August 27th, 2018, which were the biggest and smallest differences in plant height. The DEM data of maize population and canopy height were obtained with image pose correction, image mosaic, point cloud generation, and space reconstruction, et al. The canopy height and plant height were normalized, and the correlation between different cultivars and sowing date was analyzed based on UAV and manual plant height measurement. The feasibility of using DEM data of maize canopy to monitor the difference of plant height was clarified. The results showed that the height difference of maize population could be reflected by the digital elevation information obtained from high-definition RGB camera and multispectral camera. The plant height monitoring accuracy of HD RGB camera was higher than that of multispectral camera. However, the monitoring accuracy of plant height was not enough under the ready-made image equipment and treatment method. So, it was difficult to reflect the smaller plant height difference of maize population. Growth stage had a great influence on the monitoring of maize plant height. When the canopy of early growth stage has not completely covered the surface or the leaf yellow and withered in the later stage of growth. The plant height of the population affected by the exposed surface was seriously underestimated. In this study, the effects of UAV imaging equipment on monitoring maize plant height were analyzed. The influence factors can be used as reference for the application of the method in field production.
In view of the problems of constant speed and single carbon emission calculation method in the distribution model of fresh agricultural products in the transportation link of agricultural supply chain, combined with the time-varying characteristics of road network and the new multi vehicle carbon emission calculation method, this study put forward the distribution route optimization model of fresh agricultural products with four optimization objectives, which were the distribution distance, multi vehicle carbon emission, goods loss and vehicle fixed cost. In this model, the calculation of fuel consumption and carbon emission in the model would be affected by many factors, among which the load is the most important factor: Firstly, the average fuel consumption per 100 km of different trucks was calculated, then the CO2 emission factors of various trucks were calculated according to the carbon balance principle, and finally the average value of the results of each truck was taken as the carbon emission factor of the vehicle. According to those characteristics of the model, an improved double strategies co-evolutionary ant colony system (DC-ACS) was proposed. In this study, the main method was used to transform the problem into a solvable single objective problem. Then, the ant colony algorithm combined the coevolution mechanism, adaptive pheromone update strategy and local search mechanism were used to improve the solution effect of the algorithm. Finally, an appropriate fitness calculation method and stagnation avoidance strategy were designed to enhance the ability of the algorithm to jump out of local optimization. The C105 example of Solomon dataset was solved by using the improved ant colony algorithm. The optimal solutions on the four optimization objectives were 937.94 km, 4961.48 CNY, 4081.78 CNY and 7500.87 CNY respectively, which proved the effectiveness of the model proposed in this study. Based on the effectiveness of the model, the experimental results showed that the total distribution cost of the improved ant colony algorithm reduced by more than 14% on average compared with the basic ant colony algorithm on the four optimization objectives, which proved that the improved ant colony algorithm had more advantages. The improved ant colony algorithm was used to solve large-scale examples with different distributions: centralized, random and mixed. The optimal total costs were 19939.53 CNY, 24095 CNY and 24397.58 CNY, respectively. To sum up, the proposed model and algorithm could provide a good reference for the urban distribution path decision-making of cold chain logistics enterprises, a new idea to improve the distribution path optimization model and optimization method of smart agricultural supply chain, and a reference for enterprises to further expand their scale.
There are higher requirements for the timeliness of vegetable transportation and distribution. In order to solve the problems of long transportation time, high total transportation cost and short preservation time of vegetables during transportation, considering the constraints such as vehicle load and time window, this study proposed a genetic simulated annealing algorithm (GA-SA) for multi-objective vegetable distribution path optimization with time windows. That was, the simulated annealing algorithm (SA) adaptive (Metropolis) acceptance criterion was introduced into the operation process of genetic algorithm (GA). The basic idea was: First, the original population was selected, crossed and mutated by genetic algorithm to form a new generation of path population. At this time, by introducing metropolis acceptance criterion, and then, after modifying the sub situation of the new generation path population and selecting cross mutation, a new target path population was obtained. The improved algorithm retained the excellent individual, and the convergence speed, jumped out of the local optimal solution found based on genetic algorithm, and then found the global optimal solution. Then, the multi-objective of returning all vehicles to the distribution center after distribution was the least time-consuming, the lowest cost and the least use of vehicles was achieved, and the optimal path of vegetable transportation was obtained. Taking Baoding city in Hebei province as the distribution center and some towns under the jurisdiction of Baoding city as the distribution points, the experiment of vegetable transportation path optimization was designed. The experiments of genetic algorithm, simulated annealing algorithm and genetic simulated annealing algorithm were carried out, respectively. The comparative analysis was carried out from the aspects of convergence speed, total distance, total time, vehicles and total cost. The experimental results showed that, compared with the genetic algorithm and simulated annealing algorithm, GA-SA could effectively accelerate its convergence speed. The total cost of the optimized distribution route reduced by about 23.7% and 4% respectively, the total distance reduced by 22.6% and 3% respectively, the time consumption reduced by 26.2 and 2.6 hours respectively, and 2 and 1 vehicles were used less respectively. This study could also provide reference for the research of cold fresh food and other transportation path optimization.
