Intelligent excavating systems have attracted much interest because of their ability to realize cost reduction and improved productivity and safety. This study focuses on the planning of tasks and operation of intelligent excavators that autonomously operate heavy equipment. These systems have many ...
Intelligent excavating systems have attracted much interest because of their ability to realize cost reduction and improved productivity and safety. This study focuses on the planning of tasks and operation of intelligent excavators that autonomously operate heavy equipment. These systems have many applications in the construction industry. The nature of the work environment of an excavator is very hazardous when compared to that of general indoor robots. In addition, to cope with such difficult operating conditions, these systems must enable robust decision making to ensure that the system operates safely. Therefore, it is impossible to operate the intelligent excavator without the help of a well-designed task planner or algorithm that enables the system to know how to move or dig, and how to deal with unexpected situations such as obstacles. This study focuses on the task planning (maneuvering, digging, and loading) and operation of intelligent excavators. The objective of this study is to plan the task of an intelligent excavator such that the autonomous task performance reflects that of a skilled human operator's knowledge. Also, to ensure that the system operates effectively, all of the processes and protocols are defined to communicate with each of the sub-systems.
There have been many studies related to the field of intelligent excavators. Much of the research in intelligent excavation has focused on motion planning and hydraulic control systems in a single-platform state. However, only a few studies have reported a path planning strategy within the context of an intelligent excavator.
In this thesis, the characteristics of earthwork are presented to solve the problems posed in this study. First, a task plan should be established to ensure avoidance of stationary/moving obstacles when the excavator is moved on any construction site. Second, to enhance productivity, the intelligent excavator must create a work path that accounts for collaboration with the dump truck. However, the hydraulic actuation system of the excavator has limited power and is highly non-linear, making it difficult to control every motion. Therefore, the task planner should provide sophisticated plans that consider such constraints.
A complete coverage path planning (CCPP) algorithm was developed by taking account of the characteristics of earthwork and the constraints of the intelligent excavator to enable the efficient loading of dump trucks. The algorithm has a unique feature in that its cost function refers to the accessibility of the dump truck and the external condition of the work environment: this enables the collaboration with the dump truck to be maximized to reflect real-world solutions, whereas most of the CCPP algorithms concern only the moving distance and internal work environment.
To analyze the performance of this algorithm, an evaluation method that can give a quantitative result of the “Path Similarity” was developed using the “Maximum distance.” The reason for adopting the concept of the path similarity is that the goal of the algorithm must be incorporated with that of a skilled operator. Using this evaluation method, five different examples of construction sites were used to analyze the performance between the CCPP and the skilled operator’s knowledge.
Task planners have been developed to operate computerized systems using task planning algorithms. Because an important feature of a task planner is to provide visual information in real-time, visual indications of the movement of the excavator and changes in the motion (kinematic analysis) and terrain model were also presented. The virtual reality modeling language (VRML) was used to represent the model.
Further, to operate the intelligent excavator, each sub-system requires a large amount of data. The data communication manager (DCM) enables the efficient allocation of data with respect to the protocol. In this study, we define every process that can occur in the system during operation. Also, a string-type protocol was developed and is presented here in accordance with the flow of data. Because of its ease of operation, TCP/IP-based socket communication was adopted; using a dummy test and 10 field tests, we confirmed that DCM is robust in terms of the reliability of data transmission.
The task planner of the intelligent excavator has resulted in autonomous performance that approaches that of a skilled operator in terms of the path similarity. The autonomous operation achieved throughout the series of field tests shows that both the task planner and DCM are sufficiently robust.
Intelligent excavating systems have attracted much interest because of their ability to realize cost reduction and improved productivity and safety. This study focuses on the planning of tasks and operation of intelligent excavators that autonomously operate heavy equipment. These systems have many applications in the construction industry. The nature of the work environment of an excavator is very hazardous when compared to that of general indoor robots. In addition, to cope with such difficult operating conditions, these systems must enable robust decision making to ensure that the system operates safely. Therefore, it is impossible to operate the intelligent excavator without the help of a well-designed task planner or algorithm that enables the system to know how to move or dig, and how to deal with unexpected situations such as obstacles. This study focuses on the task planning (maneuvering, digging, and loading) and operation of intelligent excavators. The objective of this study is to plan the task of an intelligent excavator such that the autonomous task performance reflects that of a skilled human operator's knowledge. Also, to ensure that the system operates effectively, all of the processes and protocols are defined to communicate with each of the sub-systems.
There have been many studies related to the field of intelligent excavators. Much of the research in intelligent excavation has focused on motion planning and hydraulic control systems in a single-platform state. However, only a few studies have reported a path planning strategy within the context of an intelligent excavator.
In this thesis, the characteristics of earthwork are presented to solve the problems posed in this study. First, a task plan should be established to ensure avoidance of stationary/moving obstacles when the excavator is moved on any construction site. Second, to enhance productivity, the intelligent excavator must create a work path that accounts for collaboration with the dump truck. However, the hydraulic actuation system of the excavator has limited power and is highly non-linear, making it difficult to control every motion. Therefore, the task planner should provide sophisticated plans that consider such constraints.
A complete coverage path planning (CCPP) algorithm was developed by taking account of the characteristics of earthwork and the constraints of the intelligent excavator to enable the efficient loading of dump trucks. The algorithm has a unique feature in that its cost function refers to the accessibility of the dump truck and the external condition of the work environment: this enables the collaboration with the dump truck to be maximized to reflect real-world solutions, whereas most of the CCPP algorithms concern only the moving distance and internal work environment.
To analyze the performance of this algorithm, an evaluation method that can give a quantitative result of the “Path Similarity” was developed using the “Maximum distance.” The reason for adopting the concept of the path similarity is that the goal of the algorithm must be incorporated with that of a skilled operator. Using this evaluation method, five different examples of construction sites were used to analyze the performance between the CCPP and the skilled operator’s knowledge.
Task planners have been developed to operate computerized systems using task planning algorithms. Because an important feature of a task planner is to provide visual information in real-time, visual indications of the movement of the excavator and changes in the motion (kinematic analysis) and terrain model were also presented. The virtual reality modeling language (VRML) was used to represent the model.
Further, to operate the intelligent excavator, each sub-system requires a large amount of data. The data communication manager (DCM) enables the efficient allocation of data with respect to the protocol. In this study, we define every process that can occur in the system during operation. Also, a string-type protocol was developed and is presented here in accordance with the flow of data. Because of its ease of operation, TCP/IP-based socket communication was adopted; using a dummy test and 10 field tests, we confirmed that DCM is robust in terms of the reliability of data transmission.
The task planner of the intelligent excavator has resulted in autonomous performance that approaches that of a skilled operator in terms of the path similarity. The autonomous operation achieved throughout the series of field tests shows that both the task planner and DCM are sufficiently robust.
주제어
#토목공학
※ AI-Helper는 부적절한 답변을 할 수 있습니다.