We propose in this paper a method to match silhouettes. Silhouettes are described with a text-based representation. An iterative process is used to reduce descriptors. When the size of a little part is negligible in relation with sizes of main parts, that little part will be considered as noisy and will be suppressed from the initial textual descriptor. After the reduction process, the descriptors can be compared in order to perform the matching process.

Free

Controlling of Mean Arterial Pressure by Modified PI-ID Controller Based on Two Optimization Algorithms

Ekhlaskaram, Rawaa Haamed

Scientific article

High blood pressure is one of the diseases that most people suffer from, and it becomes a serious disease when it is not controlled precisely, especially during the surgical procedure. There must be anesthesiologists during the operation to monitor the pressure during the operation. It is not good and expensive, for patient safety and injection of the patient with the required dose, and it accurately requires an intelligent control to control the patient's pressure This paper presents nonlinear control system, to regulate the Mean Arterial Pressure (MAP) system. This controller is designed based on slate model that represent the mathematical equation that clarifies relationship between blood pressure and vasoactive drug injection. In this work Squirrel Search Algorithm (SSA) and Bacterial Foraging Optimization (BFO) are considered to optimize the controller parameters. Also nonlinear gain is used in PI-Id controller rather than fixed gain to make the controller much more sensitive to small value of error. Two algorithms applied to the controller to optimize its parameters to compare their results and determine which gives better results. The comparison results show best improvement when using the suggested controller based on SSA Algorithm. the results have no undershot with less (800s) settling time and low error.

Free

Cost Estimation of the Homogeneous Websites Using Hyper-links Analysis

Mohammed Abdullah Hassan Al-Hagery

Scientific article

Websites hide thousands of links and sub-links. Websites' links contain a huge amount of information and knowledge. This research concentrates mainly on the difficulty of early prediction of web structure size at the beginning of Website Development Life Cycle (WDLC), especially during planning and gathering requirements. There is a lack of finding an appropriate mechanism to assist developers in these steps. The objective of this research is to measure the logical size of a website in order to predict the development time and cost earlier before the development process based on based on the website contents and its internal structure. This objective includes three sub-objectives. First, analysis of seven classes of websites to collect real data sets. Second, extracting a set of relations from the gathered data and use these relations to establish a proposed model. Third, apply the gathered data in the proposed model to predict the development time and cost of a website. This research provides strong and important results that would help developers before the development process to predict total development time and cost which in turn used directly to specify development tools, draw project plan, formulate contract conditions and determine project duration and final price.

Free