An application of learning to robotic planning

Conventional approaches to robotic planning have focused on the resolution theorem prover, using general-purpose search heuristics, with the desired goal expressed in terms of logical calculus. These approaches suffer from several drawbacks; one major problem encountered in these approaches is the speed of planning. In this paper we describe an approach of applying supervised learning to robotic planning. The learning system is an intermediate one between rote learning and generalization learning, and is based on the concept of analogy. Simulation examples of various robot tasks are presented to demonstrate the significant increase in the systems's planning speed and to compare it with some existing systems.This work was supported by the National Science Foundation grant ENG-74-17586.     

Four-legged intelligent mobile autonomous robot

This paper presents the design and development of a four-legged mobile robot with intelligent sensing and decision-making capabilities. Multiple sensors with embedded knowledge bases and learning capabilities are used in a novel approach towards environmental perception and reaction. These sensors continuously monitor the environment as well as their own operating parameters. Priority is given to any one or a group of sensors based on prevailing environmental conditions. Intelligent sensing is shown to be the key towards a high degree of autonomy for a mobile robot. Nicknamed Flimar, this robot has the ability to function at varying degrees of intelligence made possible by an object-oriented architecture with embedded intelligence at various levels. This architecture is shown to be conducive towards incremental learning. Each of the four legs has three degrees of freedom, i.e. Flimar has a total of 12 motors on its four legs. Flimar can walk and turn without dragging or skidding, and also turns about its center of gravity with a zero radius. Flimar responds to light, sound and touch in different ways, based on prevailing environmental conditions. The overall goal of the paper is to present a novel walking principle and control architecture for a walking robot.     

农业智能系统的知识获取技术研究

作者及其团队长期针对农业领域的知识获取技术进行了系列性研究.阐述了运用智能引导、机器学习、数据挖掘、智能计算等技术的人工和自动/半自动的知识获取方法.这些方法能够有效地获取领域知识,发现隐含模式,进行知识精化.研发了知识获取工具.这些方法和工具反映了知识获取技术对农业信息工程所起的重要作用.     

Experimental control of a single-link flexible robot arm using energy shaping

Flexible-link robotic manipulators are mechanical devices whose control can be rather challenging, among other reasons because of their intrinsic under-actuated nature. This paper presents the application of an energy-based control design methodology (the so-called IDA-PBC, interconnection and damping assignment passivity-based control) to a single-link flexible robotic arm. It is shown that the method is well suited to handle this kind of under-actuated device not only from a theoretical viewpoint but also in practice. A Lyapunov analysis of the closed-loop system stability is given and the design performance is illustrated by means of a set of simulations and laboratory control experiments, comparing the results with those obtained using conventional control schemes for mechanical manipulators.     

Function-oriented design of a friction stir welding robot

The paper starts with the analysis on the need for intelligent manufacturing systems, and proposes our view of the solution to the problem from a mechatronic point of view. It discusses issues related to the flexible modular systems that produce modular flexible products. The paper takes a novel welding technology called friction stir welding as a case study, and investigates issues related to the design of a machine that could automate the process. It presents a function-oriented approach that could provide the guidelines in development of such a system. Along with the various alternatives of the friction stir welding machines concepts, various types of welding configurations are discussed in detail thereby providing necessary details for function-oriented conceptual design of a general-purpose robotic system. The paper, then, presents the system model for an articulated robot manipulator, as suggested by functional design, followed by simulations to investigate its suitability for such an application. The proposed solution not only ensures the achievement of requirements for the intelligent manufacturing systems, but also provides the necessary guidelines in design and development of such systems, as the concluding remarks. The paper is finalized by future developments and suggestions.     

基于神经网络的符号知识获取方法

本文基于神经网络的知识获取研究进行了综述,介绍了几种有效的模型和方法,并通过比较分析,提出了进一步开展研究工作的意见和看法。     

Acquiring,storing and utilizing process planning knowledge using neural networks

An approach is formulated for the automated acquisition of process selection and within-feature process sequencing knowledge from examples using neural networks. Network architecture, problem representation and performance issues are discussed.     

Knowledge tracing: Modeling the acquisition of procedural knowledge

This paper describes an effort to model students' changing knowledge state during skill acquisition. Students in this research are learning to write short programs with the ACT Programming Tutor (APT). APT is constructed around a production rule cognitive model of programming knowledge, called theideal student model. This model allows the tutor to solve exercises along with the student and provide assistance as necessary. As the student works, the tutor also maintains an estimate of the probability that the student has learned each of the rules in the ideal model, in a process calledknowledge tracing. The tutor presents an individualized sequence of exercises to the student based on these probability estimates until the student has mastered each rule. The programming tutor, cognitive model and learning and performance assumptions are described. A series of studies is reviewed that examine the empirical validity of knowledge tracing and has led to modifications in the process. Currently the model is quite successful in predicting test performance. Further modifications in the modeling process are discussed that may improve performance levels.     

LFKT:学习与遗忘融合的深度知识追踪模型

知识追踪任务旨在根据学生历史学习行为实时追踪学生知识水平变化,并且预测学生在未来学习表现.在学生学习过程中,学习行为与遗忘行为相互交织,学生的遗忘行为对知识追踪影响很大.为了准确建模知识追踪中学习与遗忘行为,本文提出了一个兼顾学习与遗忘行为的深度知识追踪模型LFKT.LFKT模型综合考虑了四个影响知识遗忘因素,包括学生重复学习知识点的间隔时间、重复学习知识点的次数、顺序学习间隔时间以及学生对于知识点的掌握程度.结合遗忘因素,LFKT采用深度神经网络,利用学生答题结果作为知识追踪过程中知识掌握程度的间接反馈,建模融合学习与遗忘行为的知识追踪模型.通过在真实在线教育数据集上的实验,与当前知识追踪模型相比,LFKT可以更好地追踪学生知识掌握状态,并具有较好的预测性能.     

An Experimental Evaluation of Integrating Machine Learning with Knowledge Acquisition

Machine learning and knowledge acquisition from experts have distinct capabilities that appear to complement one another. We report a study that demonstrates the integration of these approaches can both improve the accuracy of the developed knowledge base and reduce development time. In addition, we found that users expected the expert systems created through the integrated approach to have higher accuracy than those created without machine learning and rated the integrated approach less difficult to use. They also provided favorable evaluations of both the specific integrated software, a system called The Knowledge Factory, and of the general value of machine learning for knowledge acquisition.     

Evaluation of complex and dynamic safety tasks in human learning using the ACT-R and SOAR skill acquisition theories

This paper provides a human-centered analytical approach to learning dynamic and complex tasks using the Adaptive Control of Thought-Rational (ACT-R) and the State, Operator And Result (SOAR) models by comparing the task times of the model and the subjects. Twenty-one full time assembly line workers at a local computer company (14 men and 7 women) from ages 18-32 (Mean = 19.86 years, SD = 0.96 years) were randomly selected for this analysis. The task involved the placement of printed circuit board (PCB) components on the flow line of the desktop computer mother board manufacturing process. The overall timed performance of the subjects indicated that the match between the model and the subjects was good, resulting in an R² - value of 0.94. At the unit task level performance, and R² - value of 0.96 for placing the PCBs on the flow line. For tasks involving picking and searching of PCBs, the obtained R² - value was 0.76 and R² of 0.68 at the keystroke level. Findings revealed that the model already started out with a complete strategy of performing the task, whereas the human participants had to acquire additional learning information during the trials. Efforts will be made in the future to determine how the performance of the human subjects could be enhanced to meet or the same level as the model performance.     

一种基于模糊Petri网的不确定知识获取方法及其应用

本文介绍一种基于模糊Petri网获取专家系统不确定知识的方法,文中给出了模糊Petri网的知识表示、不确定推理算法及应用实例。     

Sonar-based mobile robot navigation through supervised learning on a neural net

For mobile robot navigation in an unknown and changing environment, a reactive approach is both simple to implement and fast in response. A neural net can be trained to exhibit such a behaviour. The advantage is that, it relates the desired motion directly to the sensor inputs, obviating the need of modeling and planning. In this work, a feedforward neural net is trained to output reactive motion in response to ultrasonic range inputs, with data generated artificially on the computer screen. We develop input and output representations appropriate to this problem.A purely reactive robot, being totally insensitive to context, often gets trapped in oscillations in front of a wide object. To overcome this problem, we introduce a notion of memory into the net by including context units at the input layer. We describe the mode of training for such a net and present simulated runs of a point robot under the guidance of the trained net in various situations. We also train a neural net for the navigation of a mobile robot with a finite turning radius. The results of the numerous test runs of the mobile robot under the control of the trained neural net in simulation as well as in experiments carried out in the laboratory, are reported in this paper.     

Autonomous Agents that Learn to Better Coordinate

A fundamental difficulty faced by groups of agents that work together is how to efficiently coordinate their efforts. This coordination problem is both ubiquitous and challenging, especially in environments where autonomous agents are motivated by personal goals.Previous AI research on coordination has developed techniques that allow agents to act efficiently from the outset based on common built-in knowledge or to learn to act efficiently when the agents are not autonomous. The research described in this paper builds on those efforts by developing distributed learning techniques that improve coordination among autonomous agents.The techniques presented in this work encompass agents who are heterogeneous, who do not have complete built-in common knowledge, and who cannot coordinate solely by observation. An agent learns from her experiences so that her future behavior more accurately reflects what works (or does not work) in practice. Each agent stores past successes (both planned and unplanned) in their individual casebase. Entries in a casebase are represented as coordinated procedures and are organized around learned expectations about other agents.It is a novel approach for individuals to learn procedures as a means for the group to coordinate more efficiently. Empirical results validate the utility of this approach. Whether or not the agents have initial expertise in solving coordination problems, the distributed learning of the individual agents significantly improves the overall performance of the community, including reducing planning and communication costs.     

Shaping robot behavior using principles from instrumental conditioning

Shaping by successive approximations is an important animal training technique in which behavior is gradually adjusted in response to strategically timed reinforcements. We describe a computational model of this shaping process and its implementation on a mobile robot. Innate behaviors in our model are sequences of actions and enabling conditions, and shaping is a behavior editing process realized by multiple editing mechanisms. The model replicates some fundamental phenomena associated with instrumental learning in animals, and allows an RWI B21 robot to learn several distinct tasks derived from the same innate behavior.     

智能机器人水果采摘识别系统设计

为了提高水果识别的准确性,减少周围环境对识别的影响,设计了一种用于智能水果采摘机器人的识别系统.采用随机森林来学习优化气味传感器的输出,其目的是更好地识别气味;利用摄像机采集到的原始图像计算出对应的本质图像,来对采集到的图像特别是有阴影的图像进行去阴影分析辨识.最后通过结合气味传感器的“气相”结果与摄像机的“色相”结果来评定采摘动作.实验证明该识别系统为后期果实的采摘提供了比较精确的参考.     

Model-free learning on robot kinematic chains using a nested multi-agent topology

This paper proposes a model-free learning scheme for the developmental acquisition of robot kinematic control and dexterous manipulation skills. The approach is based on a nested-hierarchical multi-agent architecture that intuitively encapsulates the topology of robot kinematic chains, where the activity of each independent degree-of-freedom (DOF) is finally mapped onto a distinct agent. Each one of those agents progressively evolves a local kinematic control strategy in a game-theoretic sense, that is, based on a partial (local) view of the whole system topology, which is incrementally updated through a recursive communication process according to the nested-hierarchical topology. Learning is thus approached not through demonstration and training but through an autonomous self-exploration process. A fuzzy reinforcement learning scheme is employed within each agent to enable efficient exploration in a continuous state–action domain. This paper constitutes in fact a proof of concept, demonstrating that global dexterous manipulation skills can indeed evolve through such a distributed iterative learning of local agent sensorimotor mappings. The main motivation behind the development of such an incremental multi-agent topology is to enhance system modularity, to facilitate extensibility to more complex problem domains and to improve robustness with respect to structural variations including unpredictable internal failures. These attributes of the proposed system are assessed in this paper through numerical experiments in different robot manipulation task scenarios, involving both single and multi-robot kinematic chains. The generalisation capacity of the learning scheme is experimentally assessed and robustness properties of the multi-agent system are also evaluated with respect to unpredictable variations in the kinematic topology. Furthermore, these numerical experiments demonstrate the scalability properties of the proposed nested-hierarchical architecture, where new agents can be recursively added in the hierarchy to encapsulate individual active DOFs. The results presented in this paper demonstrate the feasibility of such a distributed multi-agent control framework, showing that the solutions which emerge are plausible and near-optimal. Numerical efficiency and computational cost issues are also discussed.     

再励学习及其在移动机器人行为规划中的应用

再励学习（Reinforcement Learning,RL）是一种成功地结合动态编程和控制问题的机器智能方法,它将动态编程和有监督学习方法结合到机器学习系统中,通常用于解决预测和控制两类问题。提出了以矢量形式表示的评估函数,为了实现多维再励学习,用一专门的神经网络（Q网络）实现评判网络,研究其在移动机器人行为规划中的应用。     

Database management and analysis tools of machine induction

This paper surveys machine induction techniques for database management and analysis. Our premise is that machine induction facilitates an evolution from relatively unstructured data stores to efficient and correct database implementations.