Although　elastic　beam-spring　models　are　commonly　employed　to　predict　wall　deflections　in　deep　excavations,　most　existing　models　neglect　the　stiffness　hardening　behavior　of　soil.　Moreover,　the　determination　of　model　parameters　heavily　relies　on　laboratory　tests,　along　with　the　inevitable　errors　due　to　sampling　disturbance.　To　address　these　limitations,　this　study　introduces　an　enhanced　elastic　beam-spring　model　that　incorporates　non-limit　state　earth　pressure　and　soil　stiffness　hardening.　A　novel　CPTU-based　analytical　approach　is　then　proposed　to　estimate　wall　deflections　in　deep　excavations　by　correlating　in-situ　piezocone　penetration　test　(CPTU)　parameters　with　the　model　inputs.　The　validity　and　accuracy　of　the　proposed　approach　are　confirmed　through　comparison　with　previous　centrifuge　test　data.　The　proposed　approach　is　further　applied　to　a　deep　excavation　project　in　the　soft　soil　deposits　of　Taihu　Lake,　where　the　CPTU-based　predictions　show　good　agreement　with　field　measurements.　In　contrast,　solutions　derived　from　laboratory　tests　overestimate　the　wall　deflections,　likely　due　to　underestimated　soil　stiffness　caused　by　sampling　disturbance.　This　discrepancy　underscores　the　practical　advantages　of　the　proposed　CPTU-based　approach.　Finally,　the　influence　of　excavation　depth,　Young＇s　modulus　of　soil,　stiffness　of　retaining　structures,　and　strut　stiffness　on　maximum　wall　deflections　is　explored　through　the　proposed　approach.