The goal of this paper is to make a strong point for the usage of dynamical models when using reinforcement learning (RL) for feedback control of dynamical systems governed by partial differential equations (PDEs). To...
详细信息
ISBN:
(纸本)9798331540920;9783907144107
The goal of this paper is to make a strong point for the usage of dynamical models when using reinforcement learning (RL) for feedback control of dynamical systems governed by partial differential equations (PDEs). To breach the gap between the immense promises we see in RL and the applicability in complex engineering systems, the main challenges are the massive requirements in terms of the training data, as well as the lack of performance guarantees. We present a solution for the first issue using a data-driven surrogate model in the form of a convolutional Long-Short Term Memory network with actuation. We demonstrate that learning an actuated model in parallel to training the RL agent significantly reduces the total amount of required data sampled from the real system. Furthermore, we show that iteratively updating the model is of major importance to avoid biases in the RL training. Detailed ablation studies reveal the most important ingredients of the modelingprocess. We use the chaotic Kuramoto-Sivashinsky equation do demonstrate our findings.
GeoSpatial Location is often treated as the golden thread of dataanalysis while heterogeneous data is connected through the locality. Heterogeneous data integration via location is challenging and very essential in d...
详细信息
ISBN:
(纸本)9781665439022
GeoSpatial Location is often treated as the golden thread of dataanalysis while heterogeneous data is connected through the locality. Heterogeneous data integration via location is challenging and very essential in dataanalysis. For example, a rental price of a residential property is determined by interior factors (e.g., area of the property and interior design) and also external factors related to locality. The key challenge is incorporating data from various sources connected by locality and modeling the joint decision-making process. We propose to use different embeddings to collectively learn the representations of different data jointly while preserving the rental price prediction. The benefit of our heterogeneous data integration design is to model the interactions among all the factors contributing to our predefined task and achieve different levels of abstraction using multi-level spatial representations.
Based on the analysis of the working principle of the four axial filament winding machine, the 3D solid modeling of the machine was founded by the software Pro/E. The stress exerted on the turnover mechanism is consta...
详细信息
ISBN:
(纸本)9781479913909
Based on the analysis of the working principle of the four axial filament winding machine, the 3D solid modeling of the machine was founded by the software Pro/E. The stress exerted on the turnover mechanism is constantly changing due to the tension of winding fiber is constantly changing in the filament winding process. This causes the risk of exceeding the limit of the material properties. Therefore the finite element analysis on the turnover mechanism was done by the software ANSYS to verify its rationality and to ensure that the turnover mechanism can meet the operational requirements. And it can also provide theoretical basis for optimal design of the machine.
In recent years, software development has become more large-scaled, complicated, and diversified. At the same time, customer requirement of high quality and shortened delivery has increased. Therefore, we have to mana...
详细信息
ISBN:
(纸本)9780976348658
In recent years, software development has become more large-scaled, complicated, and diversified. At the same time, customer requirement of high quality and shortened delivery has increased. Therefore, we have to manage process quality and control product quality in the early-stage of software development in order to produce highly quality software products during the limited period. In this paper, we conduct multivariate linear analyses by using process monitoring data, derive effective process factors affecting the final product quality, and discuss the significant process factors with respect to software management measures of quality, cost, and delivery (QCD). Then, we discuss project management on the significant process factors affecting QCD- and show its effect on QCD.
process mining extracts relevant information on executed business processes from historical data stored in event logs. The data typically available include the activities executed, temporal information and the resourc...
详细信息
ISBN:
(纸本)9783319917047;9783319917030
process mining extracts relevant information on executed business processes from historical data stored in event logs. The data typically available include the activities executed, temporal information and the resources in charge of their execution. With such data, the functional, behavioural and organisational perspectives of a process can be discovered. Many existing process mining approaches are capable of generating representations involving the first two perspectives with all types of processes. The extraction of simple and complex resource assignment rules has also been tackled with declarative process models. However, it is noticeable that despite imperative notations like BPMN are mostly used for process modelling nowadays, the existing process mining approaches for enriching such models with resource assignments cannot discover rules like separation of duties and do not produce executable resource-aware process models. In this paper we present an approach for mining resourcea-ware imperative process models that uses an expressive resource assignment language (RALph) with the de-facto standard notation BPMN. The organisational perspective of the resulting models can be automatically analysed thanks to the formal semantics of RALph. The method has been implemented and tested with a real use case.
The development of the experimental Remote Tower Operation Human Machine Interface and the new Remote-controller work position is supported by a cognitive work and task analysis (CWA) of the presently existing work en...
详细信息
ISBN:
(纸本)9783540733188
The development of the experimental Remote Tower Operation Human Machine Interface and the new Remote-controller work position is supported by a cognitive work and task analysis (CWA) of the presently existing work environment and decision processes at airport Leipzig. This paper presents a formal approach for the description of the whole Human Machine System. It is shown how the results of a cognitive work analysis on a medium size airport are transferred into a formal executable human machine model for simulating the controllers work processes in relation to the airport processes. The model is implemented with Colored Petri Nets. The mathematical basis of Petri Nets allows a formal analysis of whole systems. Critical system states and inconsistencies in the human machine system are identified through comparison of knowledge states of the controllers with process states of the airport system by using State Space analysis. The represented formal work process model provides a valuable support for the communication between domain experts and system developers.
In the literature, there are different quality control methods for the paint curing process. In the automotive industry, direct temperature measurement and process simulation are the common methods. In the direct temp...
详细信息
ISBN:
(纸本)9781665435543
In the literature, there are different quality control methods for the paint curing process. In the automotive industry, direct temperature measurement and process simulation are the common methods. In the direct temperature measurement method, a thermograph is created with the data coming from the sensors placed on a sample vehicle while it passes through the curing process. The curing quality is determined considering the average temperature and the time elapsed above the threshold temperature value in the thermograph. Another method is simulating the paint curing process. However, these common approaches have some disadvantages. Due to the instrumentation of the sensors in the thermograph evaluation, only a certain number of vehicles can be evaluated in terms of curing quality. In addition, the differences between the locations where the sensors are placed complicate the quality analysis. In the process simulation method, it is very difficult to develop simulation models and validity of the simulation model may drastically reduce over time in the face of the variable conditions of the curing process (e.g. thermal leaks, process aging etc.). This manuscript presents a set of data-driven analytical methods of paint curing quality control.
This paper explores tractable robust optimal control of nonlinear systems with large state spaces. Conventional applications of surrogate modeling for control replace the underlying dynamical model with a data-driven ...
详细信息
ISBN:
(纸本)9781538682661
This paper explores tractable robust optimal control of nonlinear systems with large state spaces. Conventional applications of surrogate modeling for control replace the underlying dynamical model with a data-driven surrogate function. For large-scale systems, this approach possesses a host of shortcomings. We address these challenges by presenting a novel robust surrogate optimization framework for finite-time and receding horizon optimal control. Rather than modeling the entire state transition function, we define a surrogate model which maps the initial state and time series of control inputs to an approximate objective function value. We also define surrogate models which predict time series of relevant constraint functions. Since the bulk of the relevant information is encoded in the initial state, we apply a principal component analysis to project the state onto a reduced basis, allowing surrogate models with tractable parameterizations. To guarantee constraint satisfaction, we use phi-divergence to formulate distributionally robust chance constraints which are satisfied for worst-case realizations of the test datamodeling error distribution. We validate our approach using a case study of optimal lithium-ion battery fast charging using a large-scale electrochemical battery model.
The inhomogeneous Cox point process is commonly used for modeling natural disasters, such as earthquake occurrences. The inhomogeneous Cox point process is one of the popular models for the analysis of earthquake occu...
详细信息
ISBN:
(纸本)9789819904044;9789819904051
The inhomogeneous Cox point process is commonly used for modeling natural disasters, such as earthquake occurrences. The inhomogeneous Cox point process is one of the popular models for the analysis of earthquake occurrences involving geological variables. The standard two-step procedure does not however perform well when such variables exhibit high correlation. Since ridge regularization has a reputation in handling multicollinearity problems, in this study we adapt such a procedure to the spatial point process framework. In particular, we modify the two-step procedure by adding ridge regularization for parameter estimation of the Cox point process model. The estimation procedure reduces to either the Poisson-based regression or logistic-based regression. We apply our proposed method to model the earthquake distribution in Sumatra. The results show that considering ridge regularization in the model is advantageous to obtain a smaller value of the Akaike Information Criterion (AIC). Especially, Cox point process model with a logistic-based regression has the smallest AIC.
A discussion is presented of a powerful technique for establishing and maintaining control of data resources is logical datamodeling. The principal purposes of logical data models are to facilitate communications and...
详细信息
ISBN:
(纸本)0818606916
A discussion is presented of a powerful technique for establishing and maintaining control of data resources is logical datamodeling. The principal purposes of logical data models are to facilitate communications and discovery of data semantics. The IDEFlX datamodeling technique, which was developed by the US Air Force's Integrated Computer Aided Manufacturing (ICAM) program, is introduced. The author examines the evolution of IDEFlX, discusses the syntax and semantics of IDEFlX models, and introduces their roles in implementing integrated data resources.
暂无评论