For the current elections -- No. The census used to determine congressional representation is done on a decennial basis. The next decennial census will be conducted in the Spring of 2020 and released in December 2020.
As to the 2020 Census (and votes after 2020), that is harder question. Most of the people who work on the Census and will be compiling the reports are civil service employees. So if the appointed head of the Census Bureau fudges the numbers in the December 2020 repot, somebody will know the truth. Particularly if Trump loses in November 2020, it's highly unlikely that the non-political employees would keep quiet about the true results.
The bigger chance to fudge things is in how the Census Bureau conducts the census. In the past, the Census Bureau has done several things to increase the accuracy of the census. These include advertising about the census (including additional advertising in a market if returns are low in that area) and visiting addresses that fail to return a census. It also included efforts to make sure that there was an accurate count of the homeless and immigrant populations. The 2010 census actually included efforts to reach out to red states to overcome a movement on behalf of some conservatives to boycott the census.
Obviously, the Trump Administration could set priorities for advertising and follow-up that increases the chances of capturing all residents of conservative areas while skimping on resources for urban areas. That will have some marginal effect on the results. However, changing the total population figures for the state merely apportions the representatives to the states. That might benefit Republicans in the electoral college but it does not guarantee more Republican representatives in Congress. Additionally, much of the change in population is due to people moving, births, and deaths. These changes also tend to alter the partisan composition of states. The migration toward the South and West has in the past favored red states, but it has made those red states less red. (E.g., in the three of the last four cycles, Nevada has gained a seat going from one representative before the 1980 census to four representatives today. Over that same time, Nevada has gone from a lean Republican to a swing state (maybe even a lean Democrat state.) So what is a red state today might not be a red state in 2024.
The bigger thing for Congressional representation, however, is not red states versus blue states, but rather red areas within states versus blue areas within states. The census block by census block data for the states will be announced state-by-state between December 2020 and May 2021 -- with a substantial portion occurring after the winner in 2020 takes office. Again, if Trump tries to alter the top-line number for states, there would be a problem if the numbers for a given state released by a new Democratic Administration did not match the numbers for that state announced in December 2020. However, in terms of outreach by the Census Bureau, the area by area numbers are where the most mischief can occur. There are always hard to reach populations -- immigrants, homeless, people who distrust the government. By focusing the outreach efforts in rural areas and minimizing the effort in urban areas, the Census Bureau could substantially undercount urban residents. When the states then use these counts to redraw districts, this undercount would result in fewer urban districts and more rural districts -- favoring Republicans.