The teaching of computational methods in engineering education still faces problems; that is

students often find it difficult to grasp the abstract formulation of mathematics and the application of numerical algorithms in practice. This paper introduces the application of artificial intelligence in instructional design for a course on Computational Methods for undergraduates

and the focus is on the auxiliary role and effects of AI in teaching numerical computing

rather than introducing specific numerical algorithms. The three ways in which AI is used in this paper are dynamic algorithm visualisation

AI-assisted MATLAB programming guidance

and personalised error diagnosis.A typical case study of one-dimensional unsteady heat conduction will be presented to show the entire teaching process of physical model formulation and numerical discretization

algorithm implementation

result display

etc. According to the assessment data collected from classroom observations

unit tests and laboratory work

it was found that students' scores in the unit tests had increased by about 6 points

the one-time MATLAB code-passing rate had risen by 45 percentage points

and the proportion of students who could understand the algorithm in class had also increased by 33 percentage points. The above results show that AI-assisted teaching can lower students' cognitive load in numerical computation and boost their confidence and interest in programming.