in scientific research, I get confused in using terms like a model, a framework, an approach, a Method and a scheme so what is the difference between them? when I should use each of them to express my work?
In electrical engineering, a model is a (usually simplified) mathematical description of a real phenomenon. Models are developed to be able to apply mathematical tools/algorithms to real-life problems.
The term framework usually refers to a software framework (this probably applies in your case if you are in computer science). That is, a set of software tools that provide some functionality and can be extended by other programmers.
An approach is a way to tackle a problem. Once you have a mathematical model, you decide on a way to solve the model equations: this is your approach.
The term method is usually used as a synonym for algorithm. Applying your approach to your model, you design an algorithm that solves your problem, and this is your method.
The term scheme refers to a strategy, and is somewhat broader than method. For example, "quicksort" is a method that solves the specific problem of sorting a vector, but the way Stack Exchange stores the information about its users and serves it efficiently when it is needed, using (probably) a minimal amount of sorting operations might be called a scheme.
However, some of these terms are sometimes used as synonyms, in order to avoid repetition of words in articles. The same method might be alternatingly called "method", "algorithm" and "approach" in a paper, and "framework" and "scheme" might also be used interchangeably.
I can only answer this in the context of psychology, and within the context of my own understanding. A model and a framework I would interpret as more a system of how concepts are understood and explained; of the two, a framework is broader and more general, and a model usually refers to specific variables relating in a particular way to each other.
Approach and method I would use when describing a specific study or endeavour, which might be conducted within a particular framework, again going from the more general to the more specific.
For example, I might look at persuasion in the framwork of the Elaboration Likelihood Model; which although is called a model I would still consider a framework. In my study I would be interested specifically at the model of how credibility of the sender and the persuasiveness of the message are related to each other. My approach would likely be a quantitative one (rather than a qualitative), and my particular method could describe what I am actually doing in the study - perhaps a 2x2 experiment which varies two variables (sender credibility, difficulty of message).