In this research we will study an algebraic structure d-algebras. Then we establish an algorithm to check any sets in d-algebras if they are ideals or T-fuzzy d-ideals. Moreover we investigate some related properties in d-algebras.

The surveillance and reports on accidents for the current traffic management are carried out by viewing and inspecting CCTV footage. This method is slow to detect incidents and requires a lot of manpower. Since nowadays communication technology and social media play a role in reporting such incidents this research has been conducted to investigate accidents of interest from text messages reported on social media in Thai. There is currently no research or platform that supports this type of work clearly. This research presented the development of a neuron network memory-learning model to solve the problem of incidence pattern classification and incidence severity identification from social media messages in Thai. Using deep learning models such as MLP CNN Bi-LSTM and LSTM+CNN the study was designed and divided into three experimental patterns. This included examining patterns to identify traffic incidents that may be reported as general news or traffic reporting; examining patterns to indicate the type of incidents such as traffic accidents disasters damaged roads or others other than those mentioned; and examining patterns to specify the severity levels of the incidents such as normal intermediate traffic lane blocking or immovable. The experiments demonstrated the capability of CNN+LSTM learning models with the best incidence detection result at 93.44 %. The CNN model gave the best result in identifying a pattern of incidence at 85.29 % and the LSTM model was best able to determine the severity levels of the incidence at 88.53 %.

This research suggests an add-on working compatibly with Google Forms to construct multiple-choice questions (MCQs). This add-on comprises two main components including software for constructing MCQs and a data storage for collecting data used to compose MCQs. I demonstrated effectiveness of the add-on by applying it in the context of repetition online learning in which students could learn relearn and take an exam of which MCQs generated by the add-on. At the last step of this research I suggest using a sentiment analysis method – based on natural language processing and machine learning to find environmental factors and teaching strategies suitable for the repetition online learning.

Accurate efficient and robust travel time prediction is crucial to the development of advanced traveler information systems for providing route guidance information. To achieve this goal this paper proposed a travel time prediction through the matching of the current speed interval pattern to that in a historical database. Speed intervals instead of speeds are considered in this study to simplify the structure of matching patterns for improving matching efficiency. In this study speed interval patterns are defined by sets of link speed intervals that are either spatially or temporarily correlated with the link considered. With the speed interval patterns the algorithm is developed for searching the historical pattern(s) that is/are the closest match with the current one. Then link speeds from these matched patterns are combined for travel time prediction. By using the GPS probe taxi data which the collected speeds are aggregated in every 5 minutes the proposed travel time prediction system is implemented in Bangkok. With the speed data from probe taxi this paper has chosen four links/paths with different geometric and flow characteristics for testing the performance of the proposed travel time prediction system. From these tests it is found that the optimal speed interval pattern should include: 1) speeds of the studied link within three preceding time intervals and; 2) speeds of links in the first connection level of the studied link. Also while the computational time is capable of real-time application the proposed prediction algorithm is more accurate under uninterrupted flow conditions.

This thesis proposes a method for detecting harmonics current in the adaptive active harmonic filter with Recursive Least Square (RLS) algorithm. The proposed method is based on the theory of noise cancellation adaptive filter. Originally the theory was used for audio signal processing but this thesis is applied to the harmonic current detector of an active harmonic filter. The principle and construction of the harmonic current detector in the adaptive active harmonic current filter with Least Mean Square (LMS) algorithm and Recursive Least Square (RLS) algorithm are presented to compare the efficiency of both algorithms. Then simulate these two algorithms with Matlab-Simulink program to simulate the three case studies. All three case studies that simulate are generally available in industrial plants. The simulation results show that the harmonic current detector in the adaptive active harmonic current filter with the LMS algorithm and the RLS algorithm can detect the harmonic currents in the power systems. However the RLS algorithm is more efficient and has less error value than the LMS algorithm. Therefore it is confirmed that the RLS algorithm proposed in this thesis is appropriate to use with the adaptive active harmonic current filter to really work in the industry.