Research On Patent Value Assessment Of Intelligent Driving Technology Based On Support Vector Machine Improvement

With the continuous development of artificial intelligence technology,the country is strongly promoting the intelligent transformation of traditional industries and is now commonly used in many social fields.In the 14 th Five-Year Plan and the outline of the 2035 Vision,it is proposed to accelerate the deep integration of artificial intelligence with various industries and increase the core key technology research efforts.In the automotive sector,smart cars have become a global hotspot for industrial innovation,and smart driving technology is developing rapidly.With the orderly promotion of "intelligence" and "network connection",patents on related technologies have shown a global trend of continuous and rapid growth.The proportion of patents related to intelligent networked vehicles will be 19.4%in 2021,representing a year-on-year increase of 20.5%.In recent years,autonomous driving technology research and development and large-scale commercialization have landed,and the scale of investment and financing in the autonomous driving industry has continued to grow globally,with China’s autonomous driving investment and financing reaching 57.5billion yuan in 2021,an increase of 235% year-on-year.This shows that the demand for patents and investment and financing in the field of smart driving will continue to increase,however,few studies have been conducted to evaluate the patents in this field.Therefore,promoting the valuation of patents in the field of smart driving is conducive to solving the problem of patent financing,responding to the popularisation and commercialisation of smart driving technology,giving full play to the effectiveness of capital,and strongly promoting major innovations in science and technology.This paper analyses the limitations and applicability of using traditional valuation methods and the current difficulties in value assessment by analysing patents on intelligent driving technology-ACC adaptive cruise technology-and proposes the introduction of machine learning algorithms based on basic common valuation models to improve the accuracy of valuation.This paper takes the relevant data of Geely Holdings Group as a sample,qualitatively analyses the internal and external macro environment factors affecting the main business revenue,uses principal component analysis to reduce the dimensionality of the data,optimises the model parameters based on the particle swarm algorithm to construct a support vector machine model,realises the effect of the optimal combination of model parameters and predicts the main business revenue during the validity period of the patent technology under appraisal.The research results show that the prediction effect of the support vector machine model optimised based on the particle swarm algorithm is better than that of the ordinary support vector machine,effectively improving the accuracy of the model.