Numerous mobile devices have enabled Mobile Crowdsensing (MCS) to recruit a massive number of workers to sense and collect data for data requestors, thereby supporting a wide range of data-driven services. However, due to its crowdsourcing nature, MCS services face two critical challenges: ensuring privacy preservation and maintaining high-quality data collection. To address these issues, this paper proposes a privacy preserving and data quality optimization framework, named DQO-P5PI, which integrates the Preservation of Five Privacy Information (P5PI) framework with a Data Quality Optimization (DQO) scheme based on Fuzzy sets. Specifically, the P5PI frame work focuses on protecting five types of sensitive information for workers, namely the sensing data, the Reputation, the Identity, the Bid, and the Location (collectively referred to as DRIBL). In the P5PI framework, homomorphic encryption is employed to ensure DRIBL information privacy; the pseudonym mechanism is adopted to protect worker identities during interactions; Credit Verification Methods (CVM) verify the validity of workers' reputation scores, thereby protecting reputation privacy; and a Homomorphic Encryption Comparison Method (HECM) enables the platform to assess worker eligibility and select the top-k workers with the highest comprehensive scores. The DQO scheme further incorporates a Worker Reputation Update Method (WRUM) to dynamically update workers' reputation parameters, and a Credit Rating Method (CRM) that leverages Fuzzy sets to integrate bid values with historical reputation information for more reliable task allocation. We theoretically prove the correctness of the P5PI framework and DQO scheme, and extensive experimental results demonstrate that the proposed framework effectively improves sensing quality while keeping costs low.