This paper proposes a method for predicting motor vehicle emission factors by in-tegrating local driving conditions with the MOVES model.Based on second-by-second driving data collected in Xi'an,a set of driving conditions for light-duty vehicles was developed using the VSP(Vehicle Specific Power)distribution indicator.The locally developed driving conditions were then used to replace the default conditions in MOVES for micro-and meso—level case applications.The results indicate that using the default driving conditions leads to an overestimation of pollutant emis-sions at both micro and meso levels.At the meso level,the emission factors predicted using the local driving conditions are generally lower than those using the default conditions.The method established in this study enhances the portability of MOVES and improves the prediction accuracy of motor vehi-cle emissions.It also expands the application of low-time-resolution traffic data,such as sparse trajectories and average speed,in emission estimation and pollution evaluation.
In order to solve the problem of limited performance of traditional routing algorithms in software—defined networks(SDN)under large—scale and dynamically changing environments,this study proposed a routing optimization algorithm based on deep reinforcement learning.The algo-rithm realizes autonomous learning and dynamic adjustment of routing strategies by training intelli-gent agents,aiming to improve network bandwidth utilization efficiency and reduce transmission de-lay.It combines deep learning and reinforcement learning to process network status information and optimize routing decisions respectively to adapt to real—time changes in the network environment and ensure optimal performance under complex network conditions.Experimental results show that com-pared with traditional methods,this algorithm has achieved significant improvements in network throughput,latency and routing efficiency,effectively coped with the dynamics and uncertainty in SDN networks,and enhanced network performance and reliability.At the same time,the algorithm also realizes balanced allocation of network resources,further improving the overall allocation efficien-cy,which has important theoretical and practical value for building a smarter and more reliable net-work routing system.
In this paper,an improved ant colony algorithm is used to optimize constrained vehicle routing.The constraints considered include path constraints,time window constraints and capacity con-straints.The main purpose is to propose an improved ant colony algorithm to optimize vehicle routing,construct vehicle routing,and minimize the total cost of distribution routes.The ant colony algorithm is improved from three aspects:preprocessing the candidate node list involved in conditional transition probability to reduce the time complexity of route construction.An Insertion-based saving algorithm was proposed to improve the initial distribution route of ant colony to improve the optimization accura-cy.Based on ant colony system,the pheromone up-dating strategy was improved to accelerate the con-vergence rate.Based on the Solomon benchmark data set,compared with the research results obtained in recent years,the effectiveness of the improved algorithm in improving the solution accuracy and search efficiency,and the practicability in optimizing the vehicle routing problem with constraints,which expands the application field of ant colony algorithm.
Feature extraction is the key link in the process of image description, classification and analysis. Based on cognitive theory and perceptual experience, given a texture image, no matter how the texture image is transformed, rotated, scaled, affine, or even distorted, the image obtained should be the same texture from the perspective of human observation. Correspondingly, the extracted texture features should also be unchanged. Therefore, invariant texture feature extraction is an indispensable part in practical application and theoretical analysis. From the perspective of texture feature extraction, this paper studies the feature extraction and application of texture image by using the characteristics of PCNN. The icmnn model is improved by adding the connection intensity factor to couple the image structure information and weaken the gray information, so as to extract the image structure information and make the scrambling evaluation result consistent with the visual evaluation. Because the improved algorithm extracts the structure information of the image indirectly, in order to further extract the structure information directly, the image is processed by bit decomposition method, and the direct structure information is extracted by icmnn.
In order to explore how to achieve a win-win situation between China's environmental protection and technological innovation in the construction industry under the background of the current implementation of"double carbon".This paper selects China as the research object,uses the inter pro-vincial panel data from 2005 to 2019,and empirically analyzes the impact of environmental regulation on technological innovation in China's construction industry by using mixed OLS regression,random effect,fixed effect,FGLS and other estimation methods.Through the empirical study,it is concluded that there is a positive relationship between environmental regulation and the technological innovation ability of China's construction industry,and environmental regulation will promote the technological innovation of construction enterprises.Therefore,in order to promote the development of technological innovation in China's construction industry,the government should issue relevant policies to give full play to the positive incentive effect of environmental regulation on technological innovation in China's construction industry.
