Abstract: This system adopts a combination of case-based reasoning (CBR) and rule-based reasoning (RBR) to effectively simulate the thought processes of design experts, including association, intuition, analogy, induction, learning, memory, and deductive reasoning. It efficiently and coherently handles all design tasks, including scheme reasoning, equipment selection, parameter calculation, database management, panel design, and drawing and tabulation. It solves the key problems of intelligent relay protection design, enabling a more intelligent approach to problem-solving and expanding the scope of design to include all design domains, including construction drawing design. Keywords: Relay protection, case-based reasoning, intelligent design. Most practical engineering designs involve ill-conditioned structural problems. Many solutions in relay protection design cannot be described using mathematical models or even language. Therefore, it is very difficult to complete all engineering design tasks using only current rule-based reasoning (RBR) and process-based intelligent design methods. This is also the main reason why current computer intelligence is limited to the scheme design stage. However, since human experts can perform design work so perfectly, intelligent science, which is endlessly approaching human thinking, should research and implement corresponding intelligent methods. Through practical experience in power plant and substation design, it has been found that the actual design approach is quite simple and practical. Besides employing necessary rule-based reasoning methods, it leverages successful past examples (or typical designs) and then modifies and adjusts them according to the new problem to complete the new design. For example, when designing the relay protection for a new 220 kV substation, it is often necessary to modify, adjust, and supplement the design based on previous 220 kV substation relay protection designs, taking into account the characteristics of the equipment used in the new substation. This way of thinking, which directly maps the design result to the requirements of the design problem and the initial state, coincides with the recently emerging Case Based Reasoning (CBR) method in the field of intelligent systems. 1. Case Based Reasoning (CBR) Technology The CBR method is based on human cognitive processes. Its core idea is that when solving a problem, reasoning is based on successful examples of solving similar problems in the past, rather than starting from scratch. A typical case-based reasoning process can be summarized as follows: a) Retrieve the most similar old case from the case library based on the description of the current problem; b) Adjust this old case to adapt to the new problem, forming a new case-based solution for the current problem; c) Add the new case to the case library according to a certain learning strategy. It is evident that the CBR method is actually an analogical reasoning method, and analogical reasoning exists to varying degrees in various application fields. In recent years, international research on the CBR method has begun in the field of engineering design, and beneficial explorations have been made in areas such as architectural engineering and mechanical product design. It is believed that in the near future, this analogical reasoning method will be applied to various fields of power systems. 2. The Necessity of Introducing the CBR Method into the Field of Intelligent Relay Protection Design Based on the working process and characteristics of the CBR method, we can see the necessity of introducing this method into the field of intelligent relay protection design: a) Relay protection design itself is a field with a relatively weak professional theory; much of the knowledge within it is not yet well understood by humans and therefore cannot be expressed using rules, logic, mathematical models, or even general data forms. For example, existing reasoning technologies almost entirely fail to support reasoning processes based on drawing data, hindering the progress of intelligent CAD. However, examples can be represented in any form of target solution, such as data, tables, or even graphics, making the CBR method readily expressible for drawing and design knowledge. b) Some design knowledge, while ultimately expressible as rules, resides deeply within the minds of experts. Unearthing and organizing this knowledge requires significant effort and time from experts and knowledge engineers, which is precisely the "bottleneck" of the RBR method, especially in the CAD field. c) While some design knowledge is easily expressed using rules, the mapping operation rule base becomes enormous as the design process becomes more complex. This increases the workload of editing the knowledge base and inevitably slows down design reasoning. Moreover, in practical design, designers only care about the initial design state and the design result, not the intermediate reasoning steps and states. This is the so-called "expert's brainstorming, immediate solution" approach. This approach aligns perfectly with the characteristics of CBR: a reasoning process that directly maps from the initial design state to the design result. d) For pure RBR systems, not only is the editing workload large, but maintenance is also difficult. This is because there are still many dependencies between domain rules and between domain rules and control rules, making the debugging workload of adding or modifying knowledge in the system very large. Therefore, although theoretically, intelligent design can always be achieved using RBR, in the end, only a prototype system can be built, and it cannot adapt to the actual environment. e) The design work itself is actually a process of association, exploration, and learning. Each design is a learning and improvement of the previous design. However, since the RBR system itself does not learn, and its knowledge base does not expand on its own, it can only solve predefined problems. Learning is a basic function of the CBR system. It can use the reasoning results to summarize case knowledge through the learning mechanism and then add it to the case library on its own. This is also the guarantee that the intelligent design system can adapt to design changes and has vitality. f) The CBR method can at least present design cases similar to the current design to the designers for reference and modification, so that they do not have to start from scratch with those tedious and repetitive tasks that have no innovative significance. This part of the work often accounts for most or even the majority of the entire design work. In addition, the CBR system can also provide means of adjusting old cases, which of course comes from the knowledge of design experts. Although experts possess exceptional design and analysis skills, the sheer number and complexity of projects make it impossible for them to remember every adjustment detail. The CBR system, however, can accurately achieve its objectives. In short, the CBR method can effectively simulate the thought processes of experts, including association, intuition, analogy, induction, learning, and memory. However, its reasoning is often forced, unexplainable, and lacks systematicity due to its lack of deductive reasoning. Introducing the RBR method, with its strong deductive reasoning capabilities, in necessary stages (such as the scheme design phase) would significantly enhance the system's flexibility and comprehensive reasoning ability, while also noticeably reducing the burden of case retrieval and case database management. Therefore, Case and Rule Based Reasoning (CRBR) is the optimal choice for relay protection intelligent design systems. 3. Application Examples of CBR Method in Relay Protection Intelligent Design In researching and developing practical intelligent design systems, the author first introduced the CBR method into the field of relay protection design and proposed using a combination of CBR and RBR, i.e., the CRBR method, to handle computer-aided intelligent design problems. Based on the knowledge characteristics of the design process, the entire design can be divided into two stages: schematic design and construction drawing design. In the schematic design stage, the design knowledge typically comes from design specifications and design manual clauses, and can generally be expressed in heuristic rule form. Therefore, using the RBR (Reasoning Backwards) method for design reasoning in this stage is highly efficient. The specific process is as follows: input the generator-transformer unit primary system structure and parameters, start the inference engine to match rules in the rule base, find usable rules and activate them, and then generate the inference results for the schematic design. The task of the construction drawing design stage is to derive all the final construction drawing design documents from the schematic design results. However, the causal relationships in this derivation process are very complex, making it difficult to express in concrete rule form. On the other hand, designers accumulate a large number of typical design documents for various design schemes in the design library, and there is a certain mapping relationship between the schematic design results and typical design documents. These typical design documents are actually successful examples from previous designs, and this mapping relationship corresponds to case-based reasoning. Therefore, the CBR (Consciousness Backwards) method can easily describe and handle this complex causal relationship. This system employs a combination of CBR and RBR methods, effectively simulating the thought processes of design experts, including association, intuition, analogy, induction, learning, memory, and deductive reasoning. It efficiently and coherently handles all design tasks, including scheme reasoning, load statistics, equipment selection, parameter calculation, database management, panel design, and drawing and tabulation. 4. Conclusion This paper introduces a case-based reasoning method into the field of intelligent relay protection design. This method effectively mimics the analogical reasoning thinking of designers who directly map design conditions to the final construction drawings. Combined with deductive reasoning, it complements the shortcomings of both methods and solves key problems in intelligent relay protection design. This allows for a more intelligent approach to solving design problems, expanding the scope of design capabilities to include all design domains, including construction drawing design.