My homework wants me to plot all crystal structures determined per year, but when I search on PDB for all structures, many of the articles say "crystal structure," but some say "solution structure" or even "complex structure." Are all of these considered to be subsets of "crystal structure"?
Depending on who you talk to (typically laymen and scientists who are not structural biochemists), the term "crystal structure" is frequently taken to mean "some determination of molecular structure, whether by actual crystallography or some other means".
For example, a graduate student is talking to her adviser, an immunologist, about the structure-function relationship between a certain viral capsid protein and the peptide antigens they are using to create families of antibodies to the virus. The adviser consistently refers to the recently-published "crystal structure" of the protein, even though the actual study was performed using a combination of NMR spectra obtained from solution, electron microscopy, and quantum mechanical computational chemistry modeling techniques, and no crystals were actually studied.
So, you probably want to talk to your teacher/professor and clarify what s/he is looking for. Most likely, if you're in some sort of undergrad or first-year graduate class, they're just looking for "structural determination" in general, but they may specifically be looking for those models that were generated using actual crystals, so asking is always best.
I should note that there are significant methodological and computational differences when determining structure via X-ray crystallography, various types of NMR, electron microscopy (with or without crystals as the target), so-called "complex" determinations, etc. Find a structural biologist to learn much, much more.
I agree with what MattDMo has said, but just to emphasise - a 'solution' or 'NMR' structure is not a crystal structure. ('Complex structure' probably means structure of a complex. e.g between a protein and a ligand.) To be a crystal structure the structure should be determined using X-ray crystallography.
From the Protein Data Bank home page choose Advanced Search.
Then use the pull-down menu Choose a Query Type: to select Experimental Method. This will automatically populate with the X-RAY option which is what you want - have a look at the other options.
You can then use Add Search Criteria to append a new field where you can select a date range.