FA/Directory of Flight OperationsOCT 20 196767-FM-T-95FM/Deputy ChiefMore about LM radar rendezvous data processing
1. This note is in answer to some questions you scribbled on the bottom of my LM radar observations memo (67-FM-T-81). You asked if there is any smoothing process on the LM radar data prior to its input to the space- craft computer. My understanding is that there is not. That is, when the computer wants it, it reads the range/range rate register and the angle CDU's. In principle, this is the same as it was for Gemini, wasn't it? Of course, there is a check to confirm that there is tracker lock-on, the so-called “data good” signal from radar to computer.
2. Internal editing is considerably more complicated. First of all, a (side lobe) check is made to see that the radar angles measured are within 3° of those expected by the computer based on the current state vector. If either of the angles fall outside this 3° check, all of the data is re- jected, however, the crew is informed of the situation and is given the opportunity to override and force acceptance of the data. Rejection or acceptance of the data involves all four pieces of information—range, range rate, and the two angles—simultaneously, that is, you cannot se- lectively use components of the total observation individually.
3 . The next data check within the computer is based on the change in the state vectors resulting from the processing of these observations. This processing is done on each individual component of the observation-range, range rate and each of the two angles—and in each case a test is made to see if the change in either position or velocity is greater than some pre- set limit. If it exceeds thin limit, the computer stops and presents the crew the option of overriding the computer and forcing acceptance of the new solution. He must make this decision based on rather limited informa- tion and, of course, will have nowhere near the skill of our highly spe- cialized RTCC Data Select people with their magnificent computation and display capability. As the amount of radar data processed grows with time, the weight given new date is decreased substantially and I would be sur- prised if much data is rejected by this editing unless the radar goes ape.
4. As in so many other areas, it is surprising to discover how limited the analysis and justification is for the specific program formulation we have. However, when you consider the large number of different navigation pro- cesses in work I suppose we should not be surprised. An awful lot of the effort lately has been spent on estimating quality and determining how to use sextant and reticle data for rendezvous navigation, how to use VHF data, how to use telescope data for earth landmark tracking–both around the earth and moon–and, of course, there is continuing MSFN orbit deter- mination business. Each of these is a major effort and we have very few people capable of doing this work.
5. Of course, something else that increases the complexity of the LM rendezvous radar processing tremendously–and, unfortunately, in a very unresolved way right now–is the business of the rendezvous radar angle biases which apparently are an extremely complicated function of the orientation of the radar antenna off boresight. Attempts have been made to model the biases in the spacecraft computer program and estimate them in real time, along with determining state vectors. These models are almost certainly in error and actual radar performance data is very limited and of poor quality. My inclination over the past year and a half has been to eliminate this pert of the processing which forces us to do a 9X9 solu- tion as opposed to a 6×6. I'm afraid it may screw up the whole business. Unfortunately, I have not been able to prove this is a safe thing to do so far, especially from a standpoint of delta V coats.