To address the limitations of the existing Large Language Models （LLMs） in processing cross-domain knowledge， updating real-time academic information， and ensuring output quality， ScholatGPT， a scholar LLM based on Academic Social Networks （ASNs）， was proposed. In ScholatGPT， the abilities of precise semantic retrieval and dynamic knowledge update were enhanced by integrating Knowledge-Graph Augmented Generation （KGAG） and Retrieval-Augmented Generation （RAG）， and optimization and fine-tuning were used to improve the generation quality of academic text. Firstly， a scholar knowledge graph was constructed based on relational data from SCHOLAT， with LLMs employed to enrich the graph semantically. Then， a KGAG-based retrieval model was introduced， combined with RAG to realize multi-path hybrid retrieval， thereby enhancing the model’s precision in search. Finally， fine-tuning techniques were applied to optimize the model’s generation quality in academic fields. Experimental results demonstrate that ScholatGPT achieves the precision of 83.2% in academic question answering tasks， outperforming GPT-4o and AMiner AI by 69.4 and 11.5 percentage points， and performs well in all the tasks such as scholar profiling， representative work identification， and research field classification. Furthermore， ScholatGPT obtains stable and competitive results in answer relevance， coherence， and readability， achieving a good balance between specialization and readability. Additionally， ScholatGPT-based intelligent applications such as scholar think tank and academic information recommendation system improve academic resource acquisition efficiency effectively.

Drug synthesis reactions， especially asymmetric reactions， are the key components of modern pharmaceutical chemistry. Chemists have invested a lot in manpower and resources to identify various chemical reaction patterns in order to achieve efficient synthesis and asymmetric catalysis. The latest researches of quantum mechanical computing and machine learning algorithms in this field have proved the great potential of accurate virtual screening and learning the existing drug synthesis reaction data by computers. However， the existing methods only use few single-modal data， and can only use the common machine learning methods due to the limitation of not enough data. This hinders their universal application in a wider range of scenarios. Therefore， two screening models of drug synthesis reaction integrating multimodal data were proposed for virtual screening of reaction yield and enantioselectivity. At the same time， a 3D conformation descriptor based on Boltzmann distribution was also proposed to combine the 3D spatial information of molecules with quantum mechanical properties. These two multimodal data fusion models were trained and verified in two representative organic synthesis reactions （C-N cross coupling reaction and N， S-acetal formation）. The R²（R-squared） of the former is increased by more than 1 percentage point compared with those of the baseline methods in most data splitting， and the MAE（Mean Absolute Error） of the latter is decreased by more than 0.5 percentage points compared with those of the baseline methods in most data splitting. It can be seen that the models based on multimodal data fusion will bring good performance in different tasks of organic reaction screening.

In the automatic driving perception system， multi-sensor fusion is usually used to improve the reliability of the perception results. Aiming at the task of object detection in fusion perception system， a object detection method based on radar and camera fusion， namely Priori and Radar Region Proposal Network （PRRPN）， was proposed，with the aim of using radar measurement and the object detection result of the previous frame to improve the generation of region proposals in the image detection network and improve the object detection performance. Firstly， the objects detected in the previous frame with the radar points in the current frame were associated to pre-classify the radar points. Then， the pre-classified radar points were projected into the image， and the corresponding prior region proposals and radar region proposals were obtained according to the distance of the radar and Radar Cross Section （RCS） information. Finally， the regression and classification of the object bounding boxes were performed according to the region proposals. In addition， PRRPN and Region Proposal Network （RPN） were fused to carry out object detection. The newly released nuScenes dataset was adopted to test and evaluate the three detection methods. Experimental results show that， compared with RPN， the proposed PRRPN can not only detect objects faster， but also increase the average detection accuracy of small objects by 2.09 percentage points. And compared with the methods by using PRRPN and RPN alone， the method by fusing the proposed PRRPN and RPN has the average detection accuracy increased by 2.54 percentage points and 0.34 percentage points respectively.

Protocol state machine can describe the behavior of a protocol, which can help to understand the behavior logic of protocol. Oriented towards text protocols, a statistical method was firstly used to extract the semantic keyword of representative message type, and an adjacency matrix was used to describe the sequential relationship between the message types, based on which the protocol states were labeled and a state transition diagram was built. The experimental results show that the method can accurately describe the sequential relationship between the message types and abstract state machine model accurately.

To solve the problem of poor self-adaptive ability in the robot obstacle avoidance, combined with evolution thought of Genetic Algorithm (GA), an Evolutionary Operant Behavior Learning Model (EOBLM) was proposed for the mobile robot learning obstacle avoidance in unknown environment, which was based on Operant Conditioning (OC) and Adaptive Heuristic Critic (AHC) learning. The proposed model was a modified version of the AHC learning architecture. Adaptive Critic Element (ACE) network was composed of a multi-layer feedforward network and the learning was enhanced by TD(λ) algorithm and gradient descent algorithm. A tropism mechanism was designed in this stage as intrinsic motivation and it could direct the orientation of the Agent learning. Adaptive Selection Element (ASE) network was used to optimize operant behavior to achieve the best mapping from state to actor. The optimizing process has two stages. At the first stage, the information entropy got by OC learning algorithm was used as individual fitness to search the optimal individual with executing the GA learning. At the second stage, the OC learning selected the optimal operation behavior within the optimal individual and got new information entropy. The results of experiments on obstacle avoidance show that the method endows the mobile robot with the capabilities of learning obstacle avoidance actively for path planning through interaction with the environment constantly. The results were compared with the traditional AHC learning algorithm, and the proposed model had better performance on self-learning and self-adaptive abilities.

With regard to the dynamic spectrum allocation on wireless cognitive network, a dynamic Bertrand game algorithm of the channel pricing of licensed users was proposed using Bertrand equilibrium. Then, the relationship between stability of Nash equilibrium and speed parameter adjustment was analyzed. Consequently, step response function was utilized to replace the non-concussive process of game, and three-value method was proposed for getting step response parameters. The simulation results show that the proposed algorithm can obtain stable channel price when the value of speed parameter is less than 0. 04. Besides, the feasibility of using a step function to analyze the concussion game process is proved, and this method is convenient for licensed users to make real-time price and bring more economic benefits.

The conventional Partial Transmit Sequence (PTS) approaches get high computational complexity and need to transmit side information, which is difficult for hardware implementation. Concerning these problems, this paper proposed an algorithm of using m sequences as phase rotation factors and transferring them by pilot information. The m sequence can reduce the complexity of Field Programmable Gate Array (FPGA) implementation and the pilot transferring phase rotation factor need no side information. The Matlab simulation proves the algorithm is effective. Meanwhile, a Peak-to-Average Power Ratio (PAPR) suppression module was designed to be implemented on FPGA, and the results show that this module not only reduces the complexity of OFDM systems, but also works well in PAPR suppression.

A novel scheme for information hiding based on chaos digital stream was proposed, a kind of chaos digital stream which can be used to hide various information between receiver and transmitter was constructed by means of encryption and randomness of chaotic dynamic systems, as the format of stream is open only to receiver and transmitter and chaotic systems have definite equations, the scheme is safer, having large capacity of hiding. At last, an example was given to account for the scheme.