Rapid Changes In Parameters Research Articles

Developmental changes in exploration resemble stochastic optimization

Human development is often described as a ‘cooling off’ process, analogous to stochastic optimization algorithms that implement a gradual reduction in randomness over time. Yet there is ambiguity in how to interpret this analogy, due to a lack of concrete empirical comparisons. Using data from n = 281 participants ages 5 to 55, we show that cooling off does not only apply to the single dimension of randomness. Rather, human development resembles an optimization process of multiple learning parameters, for example, reward generalization, uncertainty-directed exploration and random temperature. Rapid changes in parameters occur during childhood, but these changes plateau and converge to efficient values in adulthood. We show that while the developmental trajectory of human parameters is strikingly similar to several stochastic optimization algorithms, there are important differences in convergence. None of the optimization algorithms tested were able to discover reliably better regions of the strategy space than adult participants on this task.

Concept, design, and measurement methodology of a highly dynamic thin film dryer for experimental in situ drying manipulation.

This paper describes the technical setup and validation of a thin film drying device in which the conditions such as air temperature, relative humidity, and velocity can be controlled in the range of 20-200 °C, 0-1.5ms-1, and vapor mass flows up to 1200gmin-1. For the first time, it is possible to perform in situ manipulation of the drying kinetic in a highly dynamic manner. The setup allows for a precise online determination of drying and rehydration kinetics; the recording of relevant process parameters such as air humidity, air and surface temperature; and the monitoring of optically observable material properties. The concept for rapid real-time changes in drying conditions is explained, and the constructive details are elucidated. Highly precise gravimetry by a proposed measurement methodology and simultaneously avoiding condensation during rapid parameter changes is accomplished. This is achieved by the combination of minimized thermal inertia of the system, air conditioning via carrier gas-free evaporation, and a highly responsive active insulation. The influence on the drying kinetics by temperature, humidity, and air velocity is shown with experiments on aqueous maltose solution, and the experimental precision is validated. The validation showed a high grade of accuracy regarding gravimetric determination with a maximum observed mass difference of 0.21% referring to the dried product. The dynamics of the setup under in situ changes in drying conditions is exemplified by further experiments. In addition, the presented setup for the first time enables the real time manipulation and observation of kinetically coupled processes such as crystallization behavior, morphology formation, or material degradation during drying. Therefore, it has important practical value for the development of efficient and energy-saving drying methods and products with specific, tailor-made properties.

The use of simple teeth surfaces in non-conjugate gears

Nowadays, the dominant place in the practical use of rotational motion transmission mechanisms has conjugate, and, in particular, teeth gear. Reproduction of high-order surfaces used in conjugated gears requires the use of sophisticated and expensive equipment. An alternative to such gears are approximate gears. The paper analyzes the possibility of using gears with non-conjugate engagement when using the simplest surfaces - a plane, a cylinder and their various combinations, as the working surfaces of the teeth. It is shown that with the simplest outlines of the teeth it is possible to get the high-quality transmission. The results of mathematical modeling of approximate gears are presented. A comparison of the studied gears with the involute one in terms of the cyclic error of the tooth frequency is given. The resulting values of the position error and the gear ratio are minimal and, in particular, for multi-row gears can be negligible. The insensitivity of approximate gears to the errors in wheel manufacturing and gear assembly is shown. Several gear configurations and various motion transmission configurations are provided. Prospects for the use of gears with the simplest outlines of the teeth are associated with the lack of need for complex gear-processing equipment and the possibility of manufacturing and repair in the field conditions. In addition, such transmissions are indispensable for experimental installations because of the possibility of rapid changes in parameters that do not require significant time for design and manufacture.

Design of an adaptive missile autopilot considering the boost phase using the SDRE method and neural networks

An adaptive autopilot to control a skid-to-turn missile during its boost phase is designed using the state-dependent Riccati equation (SDRE) method and neural networks (NN). To address the rapid changes in parameters during the boost phase, the translational and rotational motions of the missile are modeled with time-varying velocity and inertial parameters. The autopilot with a two-loop structure is designed to perform integrated roll-pitch-yaw control of the missile with cross-coupled dynamics; each loop has a baseline controller and an adaptive controller. The baseline controller is designed using the SDRE method for reference command tracking in a nominal environment, and the adaptive controller is designed based on NN to manage uncertainty during the boost phase. Stability analysis of the closed-loop system is performed, and the performance of the proposed autopilot is demonstrated by numerical simulation.

Short or Long Memory Estimators?

This paper contributes to the question of how to choose estimator memory length in system identification. Traditional approaches to this problem have been based on stationary models for parameter time variations. This leads to a fixed trade-off between the “size” of parameter variations and the “size” of noise. However, in practice, one often experiences non-stationary parameter behaviour. The latter problem leads to the desirability of some form of time varying estimator memory. In particular, a short memory is desirable when rapid parameter changes occur. However, a long memory is desirable when infrequent parameters change occur so as to give maximal noise discrimination. The current paper discusses these issues at a conceptual rather than theoretical level. We illustrate the ideas by reference to parameter estimation in zinc galvanizing lines.

Special issue on health care simulation

Health-related issues continue to be global challenges. Both developed and developing nations are struggling with ways to provide affordable access to health services. The threat of global epidemics including HIV, SARS and the spread of other diseases worldwide threaten to undo decades of gains in life expectancy. Poverty and political unrest breeds new diseases that are rapidly spread around the globe. At the same time the organization and delivery of health services are increasingly complex and costly. Papers in this issue demonstrate the power of simulation in addressing these challenges. Computer simulation has become state of the art in many areas and its use is still growing. Two trends can be observed. First, the conceptual scope of both simulation software and models is becoming broader and more flexible. This allows investigators to combine several modelling approaches in the same application. For example, agentbased models allow researchers to combine technical as well as behavioural aspects of systems. Second, partly as a result of the first trend, we see the use of computer simulation increasingly being applied in areas where there used to be reluctance to use this approach, e.g., in policy modelling. The first three papers address health care policy issues. Pasdirtz examines the issue of control of health care expenditures in the US. State-space models of the US health care system and the US economy were developed, estimated with time series data for the late twentieth century and analyzed using modern control theory. The results suggest that control strategies designed to contain health care costs need to be targeted at health care capital investment, prices for physicians’ services and prescription drugs. Policies directed purely at cutting costs in general or increasing efficiency are likely to have little effect. The paper by Anderson and others reports the construction of an agent-based model that has been used to study humanitarian assistance policies executed by governments and non-government organizations (NGOs) for the health and safety of refugees. An experimental design was used to study the relationships among five factors: basic needs, food and water, sanitation, medical resources and security. The simulation demonstrates the critical role of security in providing for the health and well-being of refugees. The importance of security was highlighted in the recommendations that Amnesty International made regarding the African Union Mission in Sudan (AMIS). A major strength of the model is that it allows policy makers to incorporate specific characteristics of refugees and the governmental and nongovernmental organizations that are providing humanitarian aid to the camp. Policy simulations are valuable because of their ability to facilitate training and feedback concerning potential impact of decisions. In the real world policy decisions may have serious consequences. Policy simulation models are useful because they accelerate creation of scenarios, allow rapid changes in parameters, and provide a test bed for concepts and strategies. In the present application, policies can be examined to see how refugee communities might respond to alternative courses of action and how these actions are likely to affect the health and well-being of the community. The third paper in this issue demonstrates the value of simulation as a tool to analyze policies designed to contain Health Care Manage Sci (2007) 10:309–310 DOI 10.1007/s10729-007-9031-x

Simulation tools for developing policies for complex systems: Modeling the health and safety of refugee communities

The U.S. Committee for Refugees and Immigrants estimated that there were over 33 million refugees and internally displaced persons (IDPs) in the world at the beginning of 2005. IDP/Refugee communities behave in complex ways making it difficult to make policy decisions regarding the provision of humanitarian aid and health and safety. This paper reports the construction of an agent-based model that has been used to study humanitarian assistance policies executed by governments and NGOs that provide for the health and safety of refugee communities. Agent-based modeling (ABM) was chosen because the more widely used alternatives impose unrealistic restrictions and assumptions on the system being modeled and primarily apply to aggregate data. We created intelligent agents representing institutions, organizations, individuals, infrastructure, and governments and analyzed the resulting interactions and emergent behavior using a Central Composite Design of Experiments with five factors. The resulting model allows policy makers and analysts to create scenarios, to make rapid changes in parameters, and provides a test bed for concepts and strategies. Policies can be examined to see how refugee communities might respond to alternative courses of action and how these actions are likely to affect the health and well-being of the community.

Identification of non-linear and non-stationary soil properties during the 1995 Hyogoken-nanbu earthquake

An extended Kalman filter algorithm with local iteration is presented for the identification of non-linear and non-stationary soil properties. Borehole-array strong motions were recorded at a liquefied site during the 1995 Hyogoken-nanbu earthquake. In this study, a modified Kalman filtering method in which the extended Kalman filter is iteratively used at every local time-step to track rapid parameter changes is proposed. The method is then applied to the instrumented soil layer, which is modeled by an equivalent linear model. An identification of non-linear and non-stationary soil properties was conducted successfully; and non-linear restoring force–displacement relationships including progression with time were obtained.