search

hirokawa / cssrlib-data

Sample scripts and dataset for CSSRlib
Other
27 stars 8 forks source link

The performance of PPP using CLAS is bad #30

Closed whu-dyf closed 1 month ago

whu-dyf commented 1 month ago

Dear hirokawa,

I hope this email finds you well. I've been working on testing the PPP capabilities at 18 CORS sites in Hong Kong using the CLAS data from the QZSS archive. To my surprise, I found that the PPP results are much worse than the general standard PPP performance.

Initially, I thought this could be due to the test region being outside of Japan. So, I tried running the sample data in the test_ppprtk.py script. However, the results are still the same - the PPP-RTK performance is good, but the regular PPP results are still worse than the standard PPP.

I only make some changes in the sample script to change the positioning mode from PPP-RTK to PPP:

    # ppprtk = ppprtkpos(nav, rnx.pos, 'test_ppprtkclas.log')
    ppp = pppos(nav, rnx.pos, 'test_pppclas.log')

    # cstat = cs.chk_stat()
    # if cstat:
    #     ppprtk.process(obs, cs=cs)
    if (cs.lc[0].cstat & 0xf) == 0xf:
        ppp.process(obs, cs=cs)

In theory, the orbit and clock corrections sourced from CLAS should also be usable for PPP. So, my result is confusing. Any guidance or suggestions you can provide would be greatly appreciated. Thank you in advance for your help. HKCL: PPP result using CLAS hkcl153a_clasppp

sample data: PPP-RTK result using CLAS SEPT_clas_ppprtk sample data: PPP result using CLAS SEPT_clas_ppp