For the improvement of the traditional evaluation effect of the automobile sound quality, an evaluation model of automobile sound quality is constructed based on BP neural network. The first is to introduce the basic principle of the BP neural network in detail. The second is to use the MGC parameters to establish the vehicle interior sound conversion model. The converted sound characteristic parameters are taken into the WORLD model to synthesize the new sound signals. Furthermore, the wavelet decomposition method is used to remove noise from the synthesized sound signals. Finally, a sound evaluation model based on BP neural network is established. The sound quality of automobiles can be better evaluated by carrying out the ABX test and MOS test in the field of sound conversion. For the newly synthesized sound signal and the target sound signal, it can be seen that the newly synthesized sound signal is more inclined to the target sound signal, and the sound quality is better. In addition, the sound quality is tested through loudness, roughness, sharpness, and level A in the field of sound quality evaluation. The final results show that the quality of newly synthesized sound is better, and the average errors of sound signals meet the sound standard. Therefore, the constructed sound conversion model and the sound evaluation model are feasible and effective.
As a meta-heuristic algorithm based on swarm intelligence, the WOA algorithm has few control parameters and searches for the optimal solution by encircling the prey, searching for the prey, and attacking the bubble net. During the whole process, only two internal parameters A and C are utilized for the control of the exploration and development process. BWOA is simple to implement. In the process of algorithm execution, the initial population, global exploration, and local development stages have shortcomings. Therefore, it is necessary to optimize the WOA algorithm. Based on WOA, this study conducts a high-performance computing analysis and location selection of logistics distribution center space. It is concluded that: (1) by using the combination of direct logistics distribution and hierarchical logistics distribution, the WOA algorithm optimizes the cross selection strategy, the population fitness S-LO is improved, the quality of LA is guaranteed, and the chaotic S-LO mapping eliminates inferior individuals in the population. Direct distribution is carried out for bulky goods and important distribution customers, and hierarchical logistics distribution is used for customers in intensive logistics distribution destinations. (2) WOA uses the second reverse learning, chaotic mapping, and logistic chaotic mapping to improve the location update mode. The direct distribution method is mostly used for the logistics business with short journeys, fixed distribution points, and more goods delivered at one time, and logistics enterprises do not need to store and distribute goods. The uniform ergodicity of the Tent chaotic map and logistic chaotic map is improved. Ka adaptive inertia weights are a good complement to optimize the limitations of the Ao whale algorithm. (3) The inertia weight of the levy flight behavior can play a powerful role in balancing the global exploration ability and optimization performance of the intelligent algorithm. The long-term short-distance search of HED and the long-distance jump of KVAR are combined. Variant individuals undergo vector synthesis. It reduces the construction and operation costs of logistics sites and is suitable for logistics distribution under specific conditions.
Fruit growing has played a huge role in solving food supply issues in many coutries. However, the yield and quality of fruits can be affected by various diseases, and thus timely and accurate identification of disease conditions is particularly important. Currently, using image recognition and object detection technology to diagnose fruit tree diseases has become a research hotspot in forestry informatization. Convolutional neural networks eliminate the preprocessing of manual feature selection and have high recognition performance. However, it is not easy to train due to the risk of gradient disappearance. In order to achieve better recognition effect, this research addresses the problem of applying small-scale data samples through data enhancement and transfer learning, and it optimizes the model by combining the two main attention mechanisms of SE and CBAM with ResNet50. Through experiments, it is found that the CBAM ResNet50 model has the best effect, improving the application performance of the studied model in actual scenarios.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.