Study On Real-Time Monitoring Of Feeding Quantity And On-Line Control Of Operating Parameters For Tilapia Primary Processing

China is the largest tilapia producer in the world,and the tilapia processing industry is developing rapidly.The primary processing of fish mainly includes descaling,gutting and other procedure.In China,nowadays,manual feeding descaling machine and adjusting speed based on experience knob are still adopted.There is no real-time monitoring and online control of the data about the operating parameters and material status during the fish processing.With the increasing of labor costs and the growing market demand for high-quality aquatic products,it is urgent to develop relevant equipment and automatic monitoring technology.The feeding quantity is an important data related to the stable operation of fish processing production line,affecting the speed adjustment and processing quality and productivity of processing equipment such as descaling machine.In this paper,according to the primary processing of tilapia,we developed the automatic feeding equipment and the real-time monitoring equipment of feeding quantity based on machine vision,and established an online control system of operating parameters based on multi-source data and knowledge fusion rules,which will provide technical support for automation and informatization of tilapia primary processing line.The main research contents and conclusions are as follows:1.A high-throughput tilapia conveying scheme was designed,and an automatic feeding device was constructed and optimized.Aiming at the problem of real-time monitoring of tilapia feeding quantity,a two-stage tile conveying device was designed.Both height & speed difference and height-limiting rollers of the horizontal conveyor belt were used to realize the tiling status of the tilapia in the monitoring field of view,and the adhesion and overlap of tilapia are avoided as much as possible;the optimized parameter combination obtained by orthogonal test is a horizontal conveyor belt height difference of 15 cm,a speed difference of 40 rpm,and a height-limiting roller height of34 mm.The tiling rate can reach 93.65 %.2.The image monitoring method of tilapia transportation process was studied in this paper,the monitoring device was constructed,and the LRMF algorithm was proposed to achieve the effective segmentation of the background target area of the net belt.A tilapia monitoring system based on machine vision was designed and constructed,using soft triggers to collect real-time high-throughput tilapia images during the transportation process,and obtaining high-throughput clear images through frame extraction and exposure time setting.For the grid-like image background of the conveyor belt,the small connected areas and the pre-processing problem of jagged target contours during image segmentation,through the design of LRMF algorithm that combines local thresholds,the removal of small connected areas,morphological processing and median filtering to realize the extraction of ROI for different specifications of tilapia in the net belt background.3.The feed quantity monitoring model based on machine vision was studied,and dynamic correction was implemented to realize the real-time monitoring of the feed quantity.The relationship between tilapia's image area and its weight was studied.And image processing and weighing analysis of a single fish under static conditions were carried out,and the area-weight model of tilapia was constructed.The regression analysis results showed that the correlation coefficient was 0.975.Under the conditions of actual production of high-throughput dynamic tiling and conveying,1000 tilapia raw material samples of different specifications were used to modify the model parameters using the golden section method,and a real-time monitoring model of tilapia feeding volume was established.4.The multi-source data of the production line was analyzed,combined with the experience and knowledge,a rule-based online control system for operating parameters was established.Real-time acquisition schemes for main parameters such as the operating speed of primary fish processing equipment was designed.Combined with empirical knowledge and real-time data of speed,feed quantity and descaling rate,online control rules for operational parameters of tilapia primary processing was established,and the feedback operation strategy and implementation were clarified.Finally,a real-time monitoring of feeding quantity and on-line control of operating parameters for tilapia primary processing was constructed.The test results shows that the relative error of the real-time monitoring feed quantity is 5.86 %,the system can achieve closed-loop online regulation of the operating parameters based on rules,and the processing speed exceeds the required 2000 kg/h productivity requirement.