The Proto-Tethys Ocean formed during the breakup of Rodinia supercontinent represents one of the oldest Tethyan oceans.  The Yunxian-Menghai suture zone located in western Yunnan , China is widely regarded as the site of the closure of the main Proto-Tethyan oceanic basin in southwestern Yunnan.  However , the timing of the closure of the Yunxian-Menghai Proto-Tethyan Ocean remains highly debated.  This study focuses on the mafic enclaves within the Lincang Batholith in Shuangjiang area , conducting systematic petrological , zircon U- Pb geochronological , whole-rock major and trace element , and Sr-Nd isotopic analyses.  The results reveal that the Shuangjiang enclaves are primarily composed of hornblende gabbro and gabbro-diorite.  LA-ICP-MS zircon U-Pb dating indicates an Early Silurian formation age  ( 434- 432  Ma) .  These enclaves exhibit intermediate- basic compositions , with Mg#  values ( 0. 52)  significantly  lower  than those of experimental mantle melts.  Geo- chemical characteristics show no obvious negative Eu anomalies , along with negative Ba , Nb , Zr , and Hf anom- alies and weak Sr positive anomalies.  The Sr-Nd isotopic compositions display a range from depleted to slightly enriched , with initial 87 Sr/86 Sr values of 0. 703 8 to 0. 705 6 and εNd ( t) values of - 1. 1 to 1. 2.  Integrated geo- chemical characteristics suggest that Shuangjiang enclaves are formed through fractional crystallization of partial melts derived from subduction-modified lithospheric mantle.  The emplacement of Shuangjiang enclaves indicates that the closure of Proto-Tethys Ocean occurs no earlier than the Early Silurian.

To address the complex noise characteristics inherent in measured geophysical data , the reliance of tra- ditional denoising methods on prior conditions , and the difficulty in acquiring clean labels for supervised learning , this paper introduces the Noise2Noise (N2N) unsupervised learning method.  The method aims to achieve effective denoising by  leveraging  repeated  observations or adjacent measurement points.  An N2N model is constructed based on a U-Net architecture , using multiple noisy observations as training pairs , thereby enabling network train- ing without the need for clean data.  Case studies on synthetic and field data demonstrate that , compared with tra- ditional and supervised learning denoising methods , N2N achieves superior denoising performance on both resistiv- ity and seismic data.  It not only effectively suppresses random noise in geophysical data but also handles abrupt a- nomalous data.  However , as N2N does not rely on clean labeled data , it removes not absolute noise but rather ir- regular signals ; consequently , useful signals with weak regularity may also be eliminated as noise.

The Early and Middle Devonian paleomagnetic data of South China Block remain blank.  The Devoni- an in Guangxi is a natural laboratory for comprehensive stratigraphy.  This study carries out a series of detailed experiments of petrography , rock magnetism and stepwise thermal demagnetization of natural remanent magneti- zation on red sandstone samples collected from the Lower Devonian Lianhuashan Formation in Guanyang area , Guangxi in order to identify the mineral types , contents and assemblage characteristics , formation process of the magnetic minerals , to explore the possibility of preserving sedimentary remanent magnetization.  The results re- veal that quartz-dominated detrital grains exhibit pervasive granular dissolution pores and altered grain bounda- ries( rounded , serrated , or embayed morphologies) , indicative of intense diagenetic overprinting( replacement , dissolution , recrystallization) that severely compromises depositional remanent magnetization preservation.  Rock magnetic analyses show that the principal magnetic carriers are hematite ( dominant)  and  magnetite ,  with  up- ward-increasing magnetite contents in the interval of Lianhuashan Formation.  Critically , Mesozoic chemical re- magnetization is evidenced by Late Triassic overprints carried by hematite and secondary magnetite( derived from hematite alteration)  recording  modern  viscous  remanent  magnetization ( VRM) .  Mesozoic multiple phases of magmatic-thermal disturbance  events  make  it  extremely  difficult to obtain reliable paleomagnetic signals and magnetostratigraphic frame in Guangxi devonian.  The sampling area should be kept as far away as possible from the Mesozoic igneous rocks.

Ultra-high performance concrete( UHPC) is prepared with recycled fine aggregate( RFA) in saturat- ed surface-dry state.  The influence of RFA on the macro performance , micro performance and ecological bene- fits of UHPC is analyzed.  Under the same net water-binder ratio , different proportions of RFA are designed to replace quartz sand (0 , 10% , 20% and 30%) , and the effects of RFA on the mechanical properties , workabili- ty , hydration behavior , autogenous shrinkage and other properties of UHPC are analyzed.  The results show that the saturated surface-dry RFA can improve the workability and partial mechanical properties of UHPC.  In the meantime , with the increase of RFA , the autogenous shrinkage of UHPC is effectively reduced.  On the basis of improving macro performance , the micro performance is shown as not only the increasing cumulative heat , but also a new interface transition zone( ITZ) .  The impact of RFA on the ecological benefits of UHPC is assessed by LCA.  The introduction of RFA can effectively reduce the energy consumption of UHPC and the damage to the ecological environment.  By incorporating RFA , an excellent , eco-friendly and low-carbon cement-based materi- al is prepared.

In order to build a livable city with“people-oriented and green development”, the vibrant design of urban streets is the key foundation.  Firstly , three indicators affecting the street vitality are defined from the per- spective of user psychology , 150 valid questionnaires are collected , and the reliability analysis is conducted by using Cronbach’s α .  The vitality level of the street is closely related to the three indicators by mathematical sta- tistical methods , and then the design indicator system of vibrant  street  are  established.   Secondly , the weight value of vibrant street design index is obtained after detailed analysis of the collected data by combining AHP and entropy value methods.  Finally , the Quanfeng Central Plaza in Suixi County is selected as the object to re- design the vitality  indicators ,  and  the  simulation  model  is  built  by  Analogic  software.   The  simulation  results show that the pedestrian  density  and  pedestrian flow  are  greater.   It  is  further  verified  that  the  width  of side- walks , the width of non-green belts , and the shade density are the key indicators for the vibrant street design. The design method can provide technical support for the integrated design of vibrant street.

 The expansive properties of Nanning expansive soil are improved by using marble waste powder.  The optimum dosage of marble waste powder is determined by macro test.  On this basis , the influence mechanism of marble waste powder on expansive soil expansion performance is revealed by means of micro test.  The experi- mental results show that the marble waste powder has a significant inhibitory effect on the expansion of expansive soil.  Free expansion rate , load-free expansion rate and expansion force of expansive soil decrease with the in- crease of marble waste powder content.  The best inhibition effect of marble waste powder content is 25%  on ex- pansive soil.  Combined with the limit moisture content and thermogravimetric analysis test , the change trend of the improved soil expansion with the marble waste powder content is explained from the micro perspective.  With the increase of the content of marble waste powder , the adsorption capacity of soil decreases , and the content of bound water in  soil  decreases.   Liquid  limit  and  plastic  limit  decrease  with  the decrease of  strong and  weak bound water content.   Macroscopically , the  expansion  of expansive  soil  decreases  with  the  increase  of marble waste powder content.

To investigate the deterioration characteristics of soil during the peak pre-failure process , a numerical model of remolded mudstone with different mineral compositions is constructed combining XRD analysis results based on GBM model.  The macro- and micro-evolution characteristics of remolded mudstone under different ini- tial moisture content conditions is explored and its pre-peak deterioration mechanisms of the mudstone are ana- lyzed.  The results indicate that the constructed numerical model , which accounts for different mineral composi- tions , has  been  validated  for  reliability  through comparison with experimental results.  The peak pre-failure process of the soil can be divided into three stages.  The failure modes of the samples primarily exhibit either shear failure or a combination of tensile and shear failures.  In Stage  Ⅲ , soil deterioration is most severe ; during this stage , the internal anisotropy coefficient of the soil is high , particle contact forces change abruptly , and the mechanical coordination number increases rapidly before stabilizing , with strength approaching the peak value. Microcracks primarily form in Stage Ⅲ , predominantly as shear microcracks within the clay mineral particles. This study provides a new approach for exploring the pre-peak failure deterioration process of soils with different mineral compositions under different initial moisture contents.

In order to solve the problem of unknown disturbance and uncertainty in the nonlinear manipulator sys- tem , a sliding mode controller design scheme based on deep deterministic policy gradient ( DDPG)  is proposed. Firstly , the DDPG algorithm is used to make the sliding mode surface parameters in the sliding mode control dy- namically and adaptively adjust with the joint parameters to adapt to the dynamic error and external disturbance of the system.  Secondly , the adaptive proportion control is added to the controller to improve the convergence speed and control accuracy of the system.  Finally , the two-joint manipulator is selected as the research object , and the simulation experiment is carried  out  by  MATLAB/Simulink  software.   The  results  show  that  compared  with  the sliding mode controller with fixed parameters , the position tracking response time of Joint 1 and 2 of the manipula- tor under this method is reduced by 5. 71% and 56. 25% , the speed tracking response time is reduced by 6. 89% and  38. 88% ,  respectively ,  and the average position tracking error within  0. 5 s is reduced by 5. 89%  and 40. 31% , respectively , showing a better control effect.

 In view of the instability and prolonged duration of nitrogen removal in piggery wastewater treatment system , zeolite  biofilm  suspension sphere ( ZBSS )   is  used  to design autotrophic nitrogen removal for piggery wastewater treatment.  After 27 days of acclimation , with dissolved  oxygen ( DO)  concentration which is 3. 0±0. 5 mg/L and influent nitrogen loading rate(NLR) of 0. 2 kg/ (m3   ·d) , the nitrogen ammonium removal (NAR) re- mains above 90% stably and nitrate production rate(NPR) is 0. 13 kg/ (m3   ·d) .  The ZBSS can maintain consist- ent nitritation levels.  At this time , the dominant bacterial phylum is Proteobacteria , and the relative abundance of ammonia-oxidizing bacteria is 47. 6% , and nitrite-oxidizing bacteria is not detected.  After inoculation with anam- mox sludge , DO concentration is maintained at 1. 0±0. 5 mg/L.  By coupling nitritation and Anammox into ZBSS and gradually regulating operating parameters , a ZBSS-Canon device capable of autotrophic nitrogen removal is es- tablished for the treatment of piggery wastewater.  When DO concentration is 2. 5±0. 5 mg/L and the influent NLR is 0. 86 kg/ (m3   ·d) , NRR reaches 0. 70 kg/ ( m3   · d) , and the effluent NH3 -N concentration is lower than 50 mg/L after operating for 56 days.  At this time , the relative abundance of AOB decreases , while the relative abun- dance of anammox bacteria increases.  The most dominant phylum shifts from Proteobacteria to Planctomycetota. The application of ZBSS-Canon  treatment  can  effectively  reduce  NH3 -N  as  well  as  TN  in  piggery  wastewater , thereby providing technical support for achieving stable nitrogen removal and optimizing plant remediation pond ef- ficiency.

 In order to improve the efficiency and accuracy of soil moisture estimation at the regional level , a soil moisture prediction  model  based  on  MLP neural  network is proposed and compared with radial basis  function (RBF) and back propagation(BP) neural network prediction models.  Amplitude and phase are input factors , and the soil moisture value of PBO H2 O is the learning target.  The predicted value of soil moisture is obtained through iterative training by neural network model.  The results show that the three models can meet the application re- quirements in terms of reliability and prediction accuracy.  Compared with the RBF and BP models , the root mean square error of the MLP neural network model is reduced by 12. 2% and 17. 1% , respectively , and the coefficient of determination is 0. 919.  It has higher parallelism and fault tolerance , and is more consistent with the measured data .  It can be seen that this method improves the accuracy of soil moisture inversion to a certain extent , and it is feasible and effective to invert GNSS-IR soil moisture by using MLP neural network.

Aiming at the difficulty of selecting the best parameters in regional GNSS elevation fitting model , a re- gional GNSS elevation anomaly fitting method based on WOA-BP algorithm is proposed.  Whale optimization algo- rithm (WOA) is used to optimize the weights and threshold parameters in BP neural network model. According to the characteristics of whale algorithm ,such as strong optimization ability in global search ,simple operation and few adjustment parameters ,suitable weights and threshold parameters are found for BP neural network , and the opti- mized weights and threshold parameters are brought into BP neural network model for training ,so as to improve the prediction accuracy and convergence speed of combination algorithm model fitting. Through fitting experiments on some GNSS benchmark coincidence point data in actual engineering cases ,the fitting results of WOA-BP neural network fitting method ,BP neural network and least squares support vector machine ( LSSVM) fitting method are compared. The experimental results show that the WOA-BP neural network model has stronger stability and higher prediction accuracy , and can be applied to the prediction of regional GNSS elevation fitting.

To evaluate the potential of industrial-scale pyrolyzed sludge biochar ( ISB)  as a soil amendment , the effects of ISB content( mass ratio of ISB to soil ; CK , 2. 5% , 5. 0% and 10. 0%)  on soil leaching characteristics are investigated through soil column leaching experiments.  Meanwhile , the leaching kinetics of nutrients are an- alyzed by first-order kinetic equation , double constant equation , parabolic equation and Elovich equation.  Mo- reover , the Nemero composite index method and soil heavy metal environmental capacity method are selected to evaluate the fertility level and heavy metal environmental capacity of the soil after leaching , respectively.  The results show that pH of leaching solution show a decreasing trend with the content of ISB.  The levels of nutrients such as dissolved organic carbon ( DOC) , TN , NH3 -N , TP , K , Ca , and Mg in leaching solution have a signifi- cant positive correlation with the content of ISB.  The content of ISB increases the concentrations of Zn , Fe , and Mn in the leaching solution , while the concentration of Pb shows a decreasing trend with the content of ISB. The release behavior of DOC , TN , NH3 -N , TP , K , Ca and Mg is a complex process controlled by a combina- tion of multiple  factors ,  such  as  precipitation/dissolution ,  adsorption/desorption ,  complexation/dissociation , and intra-particle diffusion.  The fertility level of the soil after leaching increases from 0. 59 to 1. 02 with the con- tent of ISB , while the level of the environmental capacity of heavy metals decreases from  1. 004 to 0. 995.   In conclusion , ISB possesses the potential to be used as a soil conditioner , and its content is recommended to be not higher than 10. 0% .

 In order to further improve the accuracy and efficiency of DEM interpolation algorithm , the linear maximum entropy model algorithm is proposed in this paper to solve the problems such as the inunique results and slow operation efficiency of the maximum entropy model interpolation method.  In this method , the probabil- ity value P representing the spatial domain variable in the maximum entropy model is taken as the expansion point to carry out Taylor expansion of the entropy function , the linear formula obtained after expansion is taken as the objective function , and the mathematical expectation and variance of the elevation value of the  sample points are used to establish equality constraints.  According to the spatial correlation of the ground objects , the maximum and minimum weight inequality constraints are set up , and the maximum entropy is used to calculate the weight coefficient of grid DEM interpolation so as to determine the elevation of the fixed point.  Linprog func- tion in MATLAB software platform is used to solve the linear maximum entropy model , and the maximum entro- py model algorithm and the inverse distance weighting method are compared and tested , the results show that compared with the other two methods , the accuracy and operation efficiency are the best , so the linear maximum entropy model algorithm has good applicability in DEM interpolation.

 The Yuechengling area , situated in the western segment of Nanling metallogenic belt , is recognized as one of the crucial Caledonian W-Sn mineralization prospective zones in South China due to its superior metal logenic geological conditions and dense distribution of mineralized occurrences.  A systematic investigation into its petrogenesis-metallogenic coupling mechanisms is indispensable for both theoretical research and achieving 
exploration breakthroughs.  This study focuses on Tongyoucun tin deposit and its ore-forming granites as key re
search subjects.  A systematic investigation is conducted , encompassing petrological analyses , zircon U-Pb and 
cassiterite U-Pb geochronology , whole-rock geochemical studies , and exploring their genetic relationships with 
tin mineralization .  Geochemical results reveal that the granites in mining area exhibit SiO2  and K2 O contents of
76. 00%-77 . 66% and 4. 78%-6 . 64% respectively , with an aluminum saturation index (ASI)  of 1. 16-1. 68
 and a differentiation index (DI)  of 88. 77-95. 10 , classifying them as the highly fractionated S-type granites 
within the strongly peraluminous high-K calc-alkaline  series.  The zircon  and  cassiterite  U -Pb  dating results 
show that the intrusion age of granite is 438. 0±10 . 0 Ma .  The tin-bearing greisen-type and fracture-zone vein
type cassiterite U-Pb ages in mining area are 429. 1±3 . 4 Ma and 413. 5±4 . 1 Ma , respectively.  These data in
dicate that both magmatic intrusion and tin mineralization events occurred during Late Caledonian period , with 
mineralization slightly postdating the granite emplacement.  The significant temporal gap between tin mineraliza
tion stages suggests the existence of two-stage tin mineralization events in Tongyoucun area , potentially linked to late-stage intrusions  related  to  tin  mineralization  at  depth.   Petrogeochemical  characteristics  indicate  that  the granites in mining area are formed during the post-collisional tectonic setting of orogenic compression-to-extension transition phase , representing a product of partial melting of ancient crust enriched in clay-rich metapelitic rocks.  Comprehensive studies reveal that Caledonian magmatic activity in Yuechengling area can be subdivided into six phases , corresponding to five W-Sn mineralization events.  Late-stage intrusions exhibit a specific metallogenic affinity with W-Sn mineralization

In order to investigate the strength of longitudinal bar on reinforced recycled aggregate concrete-filled steel tube ( R-RACFST) with 100% replacement ratio of recycled coarse aggregate under axial loading , a total of sixteen stub columns including R-RACFST with two kinds of longitudinal bar are tested , using the strength of longitudinal bar  as  the  main  variable.   Some mechanical index such as failure mode , load-displacement re- sponse , load-strain response , ductility and factual toughness are contrastively analyzed.  It is found that the dual confinement effect of steel tube and reinforcement cage can significantly weaken the brittleness of recycled aggre- gate concrete( RAC) core and improve the bearing capacity and ductility regardless of the cross-section shape ; configuring longitudinal bar with higher strength makes little  sense to  improving  overall  mechanical  behavior. The following conclusions are drawn : the mechanical performance of R-RACFST with 100% replacement ratio of recycled coarse aggregate can be guaranteed and the utilization ratio can be increased , even though the ratio of longitudinal bar is relatively low; configuring longitudinal bars with the strength close to that of steel tube is rec- ommended.

Aiming at the effect of permanent magnet temperature rise on the no-load air-gap magnetic field and output thrust of IPMSLM , the  equivalent  magnetizing  current  method  is  used  to  establish  the  no-load  air-gap magnetic field model and thrust model of the motor , and the effect of permanent magnet material properties is considered to further improve the model.  The analytical and numerical solutions of the no-load air-gap magnetic field and output thrust of the IPMSLM are compared and analyzed under different temperature rises of the perma- nent magnets , and the actual temperature rises of each part of the motor are simulated by using the electromag- netic-thermal bi-directional coupling method.  The results show that the air-gap magnetic flux density and thrust decrease with the increasing temperature of the permanent magnet , the coefficients of determination of the ana- lytical and numerical solutions of the air-gap magnetic flux density curves are greater than 0. 95 , and the maxi- mum relative error of the average thrust is 1. 70% , which verifies the accuracy of the derived model.

In order to understand the internal flow field distribution and particle classification process of a multi rotor cage horizontal vortex air classifier , FLUENT software is used to numerically simulate and analyze the flow field characteristics and classification performance of an ultrafine vertical mill classifier with six rotor cages. The reliability of the numerical model is verified by comparing simulation results such as overall pressure difference , 2  μm particles content , D50 , D97 , and D100  with actual production data of the vertical mill.  To address the prob- lems such as high impact angle between airflow and rotating cage blades , easy wear of blades , and poor particle classification performance , structural optimization simulation tests are conducted on rotating cage blades.  The results show that compared with radial straight blades , rotating cage blades with a 15° installation angle can re- duce the air intake by 24% and the overall pressure difference of the unit by 31% when producing heavy calci- um carbonate powder containing 23%  of 2  μm particles.   Meanwhile , the product maintains a  similar particle size distribution and achieves a slight rise in output.

To reconstruct the sedimentary environment of northeastern margin in  North  China  Platform during Early Ordovician , reconstruct its sedimentary paleogeographic pattern , and provide data for exploration of Paleo- zoic oil and gas resources in eastern part of North China Platform , this study focuses on Ordovician Liangjiashan Formation  in  Liujiang  Basin ,  Hebei ,  to  conduct  stratigraphic ,  petrological ,  and geochemical research.   The study reveals distinct cyclic characteristics in the carbonate rocks of Liangjiashan Formation.  Sedimentary cycles is transitioned from subtidal to peritidal and from base to top. The corresponding sedimentary facies gradually e- volve from open  continental  shelf  and  platform  margin  shoal  to  open platform and restricted platform.  These changes collectively represent a comprehensive response system to sea-level fluctuations.  Geochemical analysis indicates an upward trend in Al2 O3 , SiO2 , and TiO2  contents , while CaO and Sr contents decreases , suggesting deposition in an open shallow-sea environment characterized with high-frequency sea-level oscillations and an o- verall shallowing trend.  The sedimentary sequence reveals a sea-level rise and fall pattern consistent with the ev- olutionary trend inferred from geochemical indicators.  Liangjiashan Formation exhibits a relatively low total rare earth element  content ( average  31. 38 × 10-6) , with enrichment  in  light  rare  earth  elements  and  depletion  in heavy rare earth elements.  Negative anomalies in δEu( average 0. 58)  and δCe ( average 0. 84)  aligns with char- acteristics of shallow-sea carbonate rock deposition.  By comparaing the analysis of equal ratios Y/Ho ( average 25. 13) , MgO/CaO ( average 0. 092) , V/ ( V + Ni) ( average 0. 60 ) , Th/U ( average  3. 9 ) ,  Sr/Ba ( average 7. 88) , and MnO/TiO2 ( average 0. 29) , it can concludes that the primary deposition of Liangjiashan Formation occures in deeper part of shallow-sea continental shelf , characterized by frequent storms , high salinity , and dis- tance from the shore , within a weakly oxidizing environment.  The paleoclimate  during  deposition  is predomi- nantly warm and humid.

In order to study the local compressive performance of circular-end concrete-filled steel tubular medi- um long column with built-in stiffeners , ABAQUS is used to model and analyze , and the existing relevant re- search data are used to verify the finite element model and analyze the extended parameters.  The considered pa- rameters are slenderness ratio( λ = 20 , 40 , 60 , 80 , 100 , 120 , 140) and aspect ratio(β = 2. 0、2. 5、3. 0、3. 5、 4. 0) .  When λ < 60 , the fall trend of local compressive bearing capacity with the increase of λ is small.  When λ is from 60 to 140 , the bearing capacity under local compression shows an obvious decreasing trend with the increase of λ .  When β≤3. 5 , the local compressive bearing capacity increases with the increase of β .  When β  >4. 0 , the local compressive bearing capacity decrease with the increase of β .  The failure mode of the circular end concrete filled tubular medium long column under local compressive load is that the deformation of the end of steel pipe is large , and the concrete at the loading position of cushion block is concave.  Based on finite ele- ment simulation results in this paper , the calculation formula of local pressure bearing capacity in this case is derived.  The comparison shows that the calculated bearing capacity error is 0. 85- 1. 15 , and the calculation re- sults are better.