Abstraction level plans

Different Abstraction Level Plans. The experts in this domain can change the abstraction level of the test plan according to the amount of available information and the purpose of the plan. Below are three typical situations that require plans with different levels of abstraction ... 

The spiral model works well in any context in which the current plan must depend on the outcome of earlier work expenditure increases along the spiral each cycle includes a review of results and risks, and it drives a refinement of the goals and plans of the next cycle. An early spiral typically covers a much broader area in requirements and a much narrower piece of design the situation reverses in later cycles. Lessons learned from any spiral feed back into the most abstract level of models that are affected. There are... 

In order to have multiple representations of a plan with different abstraction levels, a hierarchical planning mechanism is necessary. The method is first to sketch a plan that is complete but vague and then gradually refine the vague parts into more detailed sub-plans until the plan consists of a complete sequence of detailed components (Z) ... 

A simplified schematic of the hierarchical planning mechanism of ES-EPA is shown in Figure 2. There are four representations of a plan with different abstraction levels Level 1, 2, 3 and 4. In this figure, the rectangular boxes represent specific tasks that can not be expanded any further, and boxes with rounded corners represent abstract components. 

On Level 1 in Figure 2, the most abstract test plan is represented by a single component which is always the Test-Plan-of-Sample. The Test-Plan-of-Sample has tens of different templates in production rule form. In this case, as the sample is some kind of water sample and the purpose is the ordinary waste water analysis, the Waste-Water-Task-Set-Rule shown in Figure 4 is applied. 

The starting point for the development of graphic 3D simulation systems was the problem of planning the use of robots and offline programming. Independent modules or simulation modules integrated into CAD system were created. To enable a simulation to be conducted, the planned or real robot system must first be generated and depicted as a model in the computer. The abstraction level of the simulation model created must be adjusted to the required imitation as detailed as necessary, as abstract as possible. Once the model has been completed, an infinite number of simulations can be carried out and modifications made. The objective is to improve the processes and eliminate the possibility of planning mistakes. 

During an early phase of the IPS development one or several concepts are developed based on the requirements which result from the IPS planning. Each concept represents an IPS on an abstract level. During an advanced development stage products and services, which were jointly described with the aid of a certain concept, are now further detailed separately. As a result, the IPS which has been offered to a particular customer is described in detail (Meier and Uhlmann 2012). 

If the nature of the problem is not readily apparent, then it might be necessary to go to the rule-based level. In this case a diagnostic rule will be applied to identify the state of the plant and an action rule used to select an appropriate response. Control will revert to the skill-based level to actually execute the required actions. More abstract functions such as situation evaluation and planning will only be required at the knowledge-based level if the problem cannot not be resolved at the rule-based level. 

Abstract. In article approaches to optimization of HHP operation as heat machine are planned. Optimization is directed on achievement of the maximal efficiencies, cold-productivities or levels of temperatures. Optimization of weight and the sizes of tubular sorbers are possible. The question of the coordination of a heat emission in hydride beds and heat exchangers is especially important. To increase efficiency of HHP it is possible, both by a choice of the best hydrides, and by optimum control in regime parameters of HHP. The mathematical modelling spent both a method enumeration of possibilities, and a regression procedure analysis, testifies to extreme behaviour of regime parameters of HHP. 

Abstract. In article approaches to optimization ofHHP operation as heat machine are planned. Optimization is directed on achievement of the maximal efficiencies, cold productivities or levels of temperatures. Optimization of weight and the sizes of... 

More than material devices and experimental phenomena, analytical instrumentation includes conceptual abstractions, in the form of an idealized plan, intended to anticipate what will happen when apparatus, specimen, and scientist are united in the laboratory. Usually occurring early in the design process, the plan provides engineers with a conceptual vision of the precise ways in which the material devices unite with the specimen, and the expected response at the phenomenal level. This plan functions as a kind of thought experiment for engineers, one in which a conceptual vision is realized materially in the performance of the experiment. 

After extensively interviewing human experts and doing experiments using an early version of the prototype system, we found two important characteristics of the test planning methods used by human experts in this domain. Firstly, experts do not always make a detailed plan. They make a plan at an appropriate level of abstraction, a rough plan for one purpose and a detailed plan for another purpose. Secondly, they do not make any plan from scratch. Experts have various kinds of "templates which they use as the starting point of their planning or as part of a plan. These two characteristics are explained in more detail below. 

On Level 2, as a result of the execution of the Waste-Water-Task-Set-Rule, the Test-Plan-of-Sample of Level 1 is expanded into a set of less abstract components Sampling, Waste-Water-Task-Set and Report. The Report component is specific and cannot be expanded any more, but, Sampling and Waste-Water-Task-Set are still abstract components on this level. Both components have multiple templates as sets of production rules, and the most appropriate ones are applied. 

On Level 4, all the abstract components are expanded into specific task components. The hierarchical planning mechanism stops on this level. Then rules for detailed modification are applied to the plan. Certain components might be added or deleted after the application of these rules. 

The characterization of operating steps, i.e., the characterization of the operators, C pj9), depends on (1) the type of operation (i.e., Boolean, integer, analytic real-valued) and (2) the level of the desired detail (abstract operation over a section of a plant, primitive operations on valves, pumps, etc.). Let us see the classes of possible models, all of which could be available and used even within the same planning problem. 

---