Research On The Rapid Detection Method Of Milk Fat Particle Siz

It is becoming a more stringent requirement for food with the development of the industry.In recent years,the per capita consumption of dairy products has increased with years,ensuring the excellent quality of milk has become particularly momentous.Among them,milk fat is not only an important component of milk,but also an important nutrient,which can directly affect the quality of milk.Therefore,the rapid detection of milk fat particles can play an important role in the production process of dairy products,check the milk quality,and further promote the application of particle size detection in food quality detection.In this paper,the commonly used particle size measurement methods are classified and introduced,and the principle and process of measuring particle size are analyzed.Among them,the light scattering method is the most widely used method in particle size measurement.Its main characteristics are wide application range,good repeatability,high precision and fast speed.In the actual measurement,in order to meet the needs of rapid detection of particle size,the total light scattering method in the light scattering method is selected.The measurement of total light scattering method mainly uses the extinction value of particles for calculation.The difficulty lies in the solution of the first kind of Fredholm equation,which is a typical inverse problem,which needs to rely on the optimization algorithm for inversion calculation.The object of the study in this paper is milk fat particles,which have a spherical shape and provide a unimodal distribution suitable for non-independent methods,which require a given distribution function.In an independent manner,this article presents genetic algorithms and an example of a search for inverse problem-solving algorithms to compensate for the fast measurement and deficiencies of previous algorithms,an optimization algorithm for hybrid immune particle tests is proposed.It has built-in immune algorithms,which is the optimization and improvement of the particle test algorithm.In this article,models of inversion particle size distributions are established for the genetic algorithm,the models search algorithm,and the immune hybrid particle swarm optimization algorithm,respectively.The simulation experiments show that the genetic algorithm has high inversion accuracy,the relative error is about 2% without noise,and the maximum relative error is about 12% with random noise,but the inversion time of the algorithm is too long,which generally takes more than ten seconds;The inversion effect of the pattern search algorithm is satisfactory without noise,but its anti-noise ability is poor.In the case of 10% random noise,the inversion result is no longer available;The improved particle swarm optimization algorithm has a shorter inversion time and greatly improves the stability and anti-noise ability.It is found that the maximum relative errors before and after noise are 1% and 8% respectively.In the non-independent mode,the particle size distribution under the R-R function distribution,N-N function distribution,and L-N function distribution are reconstructed respectively.It can be concluded that the improved particle swarm optimization algorithm has significant advantages,such as short running time,fast convergence speed,and better anti-noise performance.It is suitable for the rapid measurement of milk fat particles.