Module ACS108 Feedback No exam. Students received individual

advertisement
Module Feedback ACS108 No exam. Students received individual feedback throughout year on assignments. Module Feedback ACS123 In general, during ACS123 exam session, students have shown satisfactory performance, clear understanding of the material delivered during semester one and two. The most difficult questions for the students were finding of powers of complex numbers, radius of convergence and functions integration. In the next semester it is necessary to pay more attention to these topics in lectures and tutorials. Module Feedback ACS124 Attendance at tutorial support was relatively poor in much of semester 1.
Q1 (124 only) This was largely a bookwork essay and was done moderately well by most. Q2 Many students over‐complicated this question. The first part required the derivation of a 1st order model in time constant form from a graph and thus was relatively mechanical, but many students got thrown by the use of context variables such as temperature and did not do the obvious simple things such as estimate time constant and gain. The second part required awareness of basic engineering common sense, that is the response depends upon the external temperature in that a shift in external temperature gives a simple shift in the steady‐state temperature of the object. Again many students totally missed this point and vastly over complicated what was required. Q3 On the whole this question was done fairly well with students failing on basic understanding only. Q4 onwards 1. Disadvantages of closed loop control systems were well answered by most. Advantages of closed loop control systems however, were in general less well known. 2. The basic form of a closed control block diagram was well known by most but there were consistent small errors such as omitting the 1/s term from the integral action block and omitting the summation for proportional and integral control action. 3. The Laplace and inverse Laplace questions were well answered in general demonstrating sound knowledge of these procedures across the cohort. 4. For the question on frequency response gain and phase expressions ‐ these were in general either well answered or not attempted by some students, indicating that the material was well understood by those who had revised the material or that there were gaps in the preparation of some. 5. The Fourier analysis questions were mostly well answered although some need to check the basic reasons for using Fourier series versus Fourier transforms. Module Feedback ACS1241 Attendance at tutorial support was relatively poor in much of semester 1. Q1 Many students over‐complicated this question. The first part required the derivation of a 1st order model in time constant form from a graph and thus was relatively mechanical, but many students got thrown by the use of context variables such as temperature and did not do the obvious simple things such as estimate time constant and gain. The second part required awareness of basic engineering common sense, that is the response depends upon the external temperature in that a shift in external temperature gives a simple shift in the steady‐state temperature of the object. Again many students totally missed this point and vastly over complicated what was required. Q2 On the whole this question was done fairly well with students failing on basic understanding only. Q3 onwards 1. Disadvantages of closed loop control systems were well answered by most. Advantages of closed loop control systems however, were in general less well known. 2. The basic form of a closed control block diagram was well known by most but there were consistent small errors such as omitting the 1/s term from the integral action block and omitting the summation for proportional and integral control action. 3. The Laplace and inverse Laplace questions were well answered in general demonstrating sound knowledge of these procedures across the cohort. 4. For the question on frequency response gain and phase expressions ‐ these were in general either well answered or not attempted by some students, indicating that the material was well understood by those who had revised the material or that there were gaps in the preparation of some. 5. The Fourier analysis questions were mostly well answered although some need to check the basic reasons for using Fourier series versus Fourier transforms. Module Feedback ACS125 Overall Generally, performance on the exam was reasonably strong and similar to overall performance on the module coursework. Q1 Overall performance was satisfactory, however most students missed marks due to providing insufficient detail to their solutions. Q2 Overall performance was good on all four parts of the question. In part (a) some students did not recognise the null event (represented by the epsilon symbol), which was a valid event sequence for the automaton. In part (b) some students performed a two‐tailed test rather than the requested one‐tailed test. In part (c) some students incorrectly reversed the powers‐of‐z when taking the z‐transfer function of the difference equation. Q3 The majority of the students attempted this optional question, and most students performed very well. Some students neglected the fact that the ‘increment’ solution was presented, and not the full one, however the technical approach used was mostly correct. Q4 This question was attempted by a minority of students; but those who did attempt the question tended to perform reasonably strongly. Some students simulated the system, but neglected to calculate costs and QALYs. Some students also did not provide three realistic recommendations for how the model could be improved. Module Feedback ACS126 Q1 a) out of range checker program: ∙ Many people got full marks. ∙ Several people made the mistake of thinking the program indexed the array contents from 1, whereas it indexed from 0 instead. ∙ A couple of people misinterpreted the printf statement (e.g. “value 8000 is out of valid range” was not the correct answer. ∙ A couple of people thought the program would say that all values were out of range. ∙ One person just discussed the program instead of telling me what would be printed, as requested in the question. ∙ Some people used the wrong format but the right value, and lost 1 mark. Q1 b) program to correspond to flowchart: ∙ Common errors were: o Setting MAXSIZE in an assignment statement – the flowchart doesn’t show this. o Not spotting that this was a nested loop. o Not declaring a 2D array and assigning values to it. o Not following the loops and branches of the flowchart. o Not printing a new line for each row. ∙ A few people got full marks, but overall performance was poor on this question. Q1 c) divide by 2 by shifting right ∙ Common errors were:
o Not spotting that this would result in an integer divide so there would be no fractional values. o Not realising that shift right by 1 would result in division by 2. o Misinterpreting the format specifier in the printf statement. ∙ A few people got full marks. Q1 d) calloc and char arithmetic ∙ There were many full marks. ∙ The program used calloc and char arithmetic to create a string with the English alphabet in it. You didn’t need the ascii table to solve the problem, just knowledge of char arithmetic should have sufficed, along with a knowledge that the letters are contiguous in the ascii table. ∙ The string of letters would have been printed all on the same line, but I allowed 1 char per line as an answer. ∙ Quite a few people gave answers like ‘a’+0, ‘a’+1, etc. This didn’t get any marks. Q2 a) structs for matrix and vector ∙ You should have used #define for the MAXSIZE, but I allowed the constant 100 instead. ∙ You needed the correct syntax for the struct. ∙ The matrix (in the struct) needed to be a 2D array, and it could be float, int or double. ∙ The vector (in the struct) needed to be a 1D array, same range of types allowed. ∙ The rows and columns for the matrix, and the size of the vector needed to be int, and these also needed to be in the struct. Q2 b) function prototype ∙ Many people seem to have completely forgotten what a function prototype is, but some people got this right. ∙ Common errors were: o Passing matrix and vector as parameters, but with no variable name in the prototype parameter list. o Passing specific struct members, rather than the struct itself, as parameters.
o Passing matrix and vector (to multiply) by address – they don’t need to be altered in the function, so they should be passed by value. However, I allowed this if everything else was right. o Not passing the result vector struct by address, although I did allow returning the result vector in an array, if it was done correctly. o Not using a struct at all. ∙ A few people got full marks, but overall performance was poor. Q2 c) function ∙ Common errors were: o Not using members of struct correctly. o Not derefing correctly. o Not setting the size of the output vector, which should have been a member of the vector struct. o Some errors in accumulating the sum of the products – many forgot to zero appropriately, and some forgot to accumulate at all. o Using the wrong index for the vector. ∙ There were a small number of high marks on this question. Q3 a) inputs and outputs for the system in the diagram of the assembly line ∙ Many people got full marks.
∙ The clues were in the diagram (I0.1‐I0.4 and Q0.0‐Q0.3). I allowed some marks for describing the sensors in general terms; for full marks a detailed list was necessary. Q3 b) flowchart for assembly line algorithm ∙ Common errors were: o Not looping back to show continued sampling after reject or after completed item. o Not showing branching resulting from each individual sensor reading. o Even for those with good answers, there was still a problem with managing the reject process vs. letting the assembled item drop off the belt and looping back to sample continued stream of items. ∙ There were some high marks, and 1 person got full marks. ∙ The best solutions were very detailed and showed testing each input, followed by branching to the appropriate output action. 
Download