Ancient building full life cycle dynamic management method and system based on digitization and intelligent means

Abstract

The application discloses a full life cycle dynamic management method and system for ancient buildings based on digital and intelligent means, and particularly relates to the technical field of data processing for digital protection and repair management of ancient buildings, and comprises the following steps: obtaining an ancient building digital twin base map and a pre-repair site operation record, converting a hidden part number into a structure node address by using a building part code, converting a disposal action in a work task into a disposal action bit, and outputting a hidden part twin object string; based on a target hidden part in the hidden part twin object string, the structure node address is used to extrapolate to an observed component node along a building structure edge; by locating the hidden part of the ancient building as a structure node in the digital twin base map, a zero quantifier action signature is derived based on a candidate hidden state, an action difference word is generated, and when the action difference exists, a forensic path is matched by using a forensic binary graph, so that the hidden part forensics is driven from data loss to protection management action difference.

Description

Technical Field

[0001] This invention relates to the field of data processing technology for the digital protection and restoration management of ancient buildings. More specifically, this invention relates to a dynamic management method and system for the entire life cycle of ancient buildings based on digital and intelligent means. Background Technology

[0002] In the digital protection and restoration management of ancient buildings, existing processes mainly revolve around three-dimensional mapping and archiving, disease inspection records, environmental monitoring access, and restoration data collection. A digital twin model of the building is formed through point cloud scanning, image acquisition, component annotation, sensor readings, and engineering records. However, during the pre-renovation assessment of ancient wooden buildings, the beam ends entering the wall, the roof base, the back of the brackets, the wall interlayer, the inside of the platform, and the area under the painted covering layer are often directly related to the determination of the scope of support, sealing, inspection, and renovation. Meanwhile, protection and management require minimizing dismantling, sampling, opening, and surface disturbance. In such scenarios, if the goal is to complete the model data and then supplement the measurement of hidden parts, it is easy to repeatedly remove tiles, open holes, remove decorations or expand the closed area in order to confirm a single point of doubt, which increases the disturbance to the ancient building itself. If only missing items are registered because the hidden parts are difficult to observe, the same hidden risks may be delayed for a long time before repair, additional investigations may be carried out temporarily before construction, and different people may give different opinions on the same part. The root cause is that the existing digital twin management relies more on "whether data is obtained" to arrange supplementary testing or leave blanks, without first judging whether the different states of the hidden parts are sufficient to change the protection and management actions. The technical problem this application aims to solve is: how to enable digital twins to determine whether hidden parts of ancient buildings need to be documented and the document collection path based on differences in protection and management actions, under the conditions that it is difficult to obtain sufficient evidence and the document collection process itself may cause protection disturbance, thereby reducing meaningless document collection and supporting pre-renovation decision-making. Summary of the Invention

[0003] To overcome the aforementioned deficiencies of the prior art, embodiments of the present invention provide a dynamic management method and system for the entire life cycle of ancient buildings based on digital and intelligent means. By positioning the hidden parts of ancient buildings as construction nodes in a digital twin base map, deriving wordless action signatures based on candidate hidden states and generating action difference words, and then using the evidence collection bipartite graph to match the evidence collection path when action differences exist, the evidence collection of hidden parts is transformed from being driven by data deficiency to being driven by protection and management action differences, thereby solving the problems mentioned in the background art.

[0004] To achieve the above objectives, the present invention provides the following technical solution: a dynamic management method for the entire life cycle of ancient buildings based on digital and intelligent means, comprising: S1. Obtain the digital twin base map of the ancient building and the on-site work record before the renovation. Convert the hidden part number into the construction node address using the building part code, and convert the handling actions in the work task into handling action bits. Output the hidden part twin object string. S2. Based on the target hidden part in the twin object string of hidden parts, push the construction node address outward along the building construction edge to the observation component node, write the on-site state of the observation component node, the recent intervention bit and the construction layer transmission bit into the state base segment by segment, perform binary expansion on the unobserved state bits, and output the candidate hidden state string. S3. Write the candidate hidden state into the free integer bits through the candidate hidden state string, and connect the action bit through the action grid and the entry grid to form the existence quantifier bit. Then, the Cooper Pressburg arithmetic quantifier elimination is performed by the connector, order, and remainder items to output the quantifier-free action signature. S4. Generate action difference words by bitwise XORing the action signature without quantifiers. When all bits of the action difference word are zero, output no evidence collection management record. When the action difference word contains non-zero bits, construct an evidence collection bipartite graph with the entry bit and non-zero bits, and perform Hungarian bipartite graph matching with the entry cost word, and output evidence collection path record. S5. Based on the evidence collection path record, form the evidence collection result record, perform position matching between the evidence collection result record and the candidate hidden state string, and if a match is found, write back the twin state of the target hidden part and the maintenance or repair management record of the building component product; if a match is not found, write the review twin state and output the pre-repair management decision record.

[0005] In a preferred embodiment, S1 includes: S11. Read the building hierarchy tree from the digital twin base map of ancient buildings, write the child node into the current level sequence according to the edge order from the parent node to the child node, use the binary code length of the number of nodes in the current level as the bit width of the current level, shift the address of the parent node to the left of the current level bit width and then connect it to the current level sequence, and output the building part coding table. S12. Read the hidden part number from the site work record before the repair, perform high-level segmentation on the hidden part number according to the level bit width in the building part coding table, and match the child node order in the building hierarchy tree with the segmented position value in sequence. When a single child node is hit, read the child node address as the construction node address. When there is no hit or multiple hits, write the placement verification record. S13. Read the operation action name from the pre-renovation site operation record, generate the disposal action order based on the row order of the operation action name in the ancient building renovation operation sequence table, shift the disposal action order to the left of the construction node address width and then connect it to the construction node address to form the disposal action bit, and connect the disposal action bit to the end of the construction node address, and output the hidden part twin object string.

[0006] In a preferred embodiment, S2 includes: S21. Extract the construction node address of the target hidden part from the twin object string of the hidden part. Use the construction node address as the current node address. Read the building construction edge in the digital twin base map of the ancient building. The starting address is consistent with the current node address and the direction point is to the observation end. Write the end address of the building construction edge into the extrapolation path position. If the end address does not have the field status position, replace the current node address with the end address and continue reading. If the end address has the field status position, stop reading and output the observation component node address. S22. Based on the node address of the observed component, read the on-site status bit, and in the repair time sequence table, read the work items that the construction node address hit in reverse order of the work time. Shift the work order of the hit work items to the left by the work type bit width and then connect them to the work type bit to generate the recent intervention bit and output the observation acceptance record.

[0007] In a preferred embodiment, S2 further includes: S23. Read the structural layer number of the target hidden part and the structural layer number of the observation component node. Subtract the structural layer number of the observation component node from the structural layer number of the target hidden part to generate a layer difference value. Connect the sign bit of the layer difference value to the left of the absolute value bit of the layer difference value to generate a structural layer transfer bit. Write the field status bit, the near intervention bit, and the structural layer transfer bit into the status base in bit width order. S24. Count the number of unobserved state bits with empty values ​​in the state base and generate an unobserved count. Generate a binary expansion number starting from zero and incrementing by one until the binary expansion number reaches two to the power of the unobserved count minus one. Fill each binary expansion number into the unobserved state bit and output the candidate hidden state string.

[0008] In a preferred embodiment, S3 includes: S31. Based on the candidate hidden state string, read the candidate hidden state according to the candidate sequence number, write the candidate sequence number to the front of the free integer position, write the candidate hidden state to the back of the free integer position, shift the action position obtained by converting the candidate hidden state to the left by the width of the building cell and then connect it to the action cell, shift it to the left by the width of the entry cell and then connect it to the entry cell, and output the existence quantifier position. S32. Using the address of the node constructed from the hidden part of the target as the starting address, read the edge number of the active cell along the construction edge of the ancient building digital twin base map, accumulate the edge number to form a connection item, subtract the active cell from the entering cell to generate an order item, divide the action bit by the width of the building cell to generate a remainder item, and write the connection item, order item and remainder item into the Pressburg constraint string.

[0009] In a preferred embodiment, S3 further includes: S33. When performing Cooper Presburg arithmetic quantifier elimination on the Presburg constraint string, first rewrite the linear terms containing existential quantifiers as a unified term with an existential quantifier coefficient of one, then expand the remainder terms into remainder branches, then generate the lower bound of existential quantifiers with the connecting terms, generate the upper bound of existential quantifiers with the order terms, and replace the existential quantifiers with the lower bound plus the remainder branches. S34. After each remainder branch completes the replacement, delete the existing quantifier position, count the number of remaining existing quantifier positions, output the string without quantifier constraint when the number of remaining existing quantifier positions is zero, and return to S33 to process the next existing quantifier position when the number of remaining existing quantifier positions is not zero. S35. Substitute the candidate hidden state string back into the unqualified constraint string to generate the back-substitute action bit. Perform a bitwise XOR operation between the back-substitute action bit and the disposal action bit. When all bits of the XOR result are zero, compress the unqualified constraint string into an unqualified action signature. When the XOR result contains non-zero bits, write the candidate hidden state into the conflict elimination record.

[0010] In a preferred embodiment, S4 includes: S41. Based on the action signature without quantifiers, read the action signature pair in order of candidate hidden state, perform bitwise XOR on the action signature pair to generate candidate difference words, and merge the candidate difference words into the action difference word by bitwise OR. When all bits of the action difference word are zero, output no evidence collection management record. S42. When the action difference word contains a non-zero position, read the entrance position leading to the hidden part of the target from the digital twin base map of the ancient building as the left vertex, take the non-zero position in the action difference word as the right vertex, and write the bipartite graph edge when the building construction path from the entrance position to the hidden part of the target covers the non-zero position associated action grid. S43. For the bipartite graph edge, read from the entrance position along the building construction path to the target hidden part, write the component count data into the front part of the entrance cost word, shift the contact surface count data to the left and connect it to the entrance cost word, shift the uncover count data to the left and connect it to the entrance cost word, shift the enclosed space count data to the left and connect it to the entrance cost word, and convert the entrance cost word into a matching value.

[0011] In a preferred embodiment, S4 further includes: S44. When performing label initialization through Hungarian bipartite graph matching, the left vertex label is generated by reading the previous position in ascending order of the left vertex association matching cost, the right vertex label is written to zero, and the relaxation word is generated by subtracting the left and right vertex labels from the matching cost. The bipartite graph edges with a relaxation word of zero are written to the zero relaxation edge set. S45. Using the unmatched left vertex as the root node of the alternating tree, read the right vertex along the zero-relaxed edge set. When an unmatched right vertex is read, perform a matching flip along the path of the alternating tree. When a matched right vertex is read, connect the left vertex connected to the matched right vertex to the alternating tree. S46. When no unmatched right vertex is read in the alternating tree, read the relaxation words from the left vertex inside the alternating tree to the right vertex outside the alternating tree. Read the previous relaxation words in ascending order as the label rewriting words. Add the label rewriting words to the left vertex label and subtract the label rewriting words from the right vertex label. Then return to the zero relaxation edge set generation process until the number of matched right vertices reaches the number of non-zero bits in the action difference words. Output the evidence path record.

[0012] In a preferred embodiment, S5 includes: S51. Read the target hidden part address and path sequence according to the evidence collection path record, retrieve the evidence collection operation backfill item according to the path sequence, perform right truncation or left zero padding on the original observation code in the evidence collection operation backfill item according to the candidate hidden state bit width, generate the evidence collection result status word, and append the evidence collection result status word to the end of the target hidden part address, and output the evidence collection result record. S52. For the evidence collection result status word in the evidence collection result record, read the candidate hidden state in the candidate hidden state string in incrementing the candidate sequence number, perform bitwise XOR between the evidence collection result status word and the candidate hidden state, and then perform full bit OR operation on the bitwise XOR result. When the full bit OR result is zero, write the hit candidate sequence number and stop reading. When the full bit OR result is not zero, continue reading the next candidate hidden state. When the candidate hidden state reading ends and no hit candidate sequence number is written, output the miss flag. S53. Based on the hit candidate sequence number, read the hidden status of the hit candidate, locate the twin status slot in the digital twin base map of the ancient building with the target hidden part address, rewrite the status bit to be verified in the twin status slot to the hidden status of the hit candidate, and connect the evidence collection path record, evidence collection result status word and hit candidate sequence number to the building component product maintenance or repair management record according to the path order; based on the missing mark, rewrite the status bit to be verified in the twin status slot to the verification twin status, and output the pre-repair management decision record.

[0013] A dynamic management system for the entire lifecycle of ancient buildings based on digital and intelligent technologies. The system includes a twin positioning module, a state expansion module, a quantifier elimination module, a path matching module, and a decision writing module. The twin positioning module is used to obtain the digital twin base map of the ancient building and the on-site work record before the renovation. It converts the hidden part number into the construction node address by the building part code, and converts the disposal actions in the work task into disposal action bits, and outputs the hidden part twin object string. The state expansion module, based on the target hidden part in the hidden part twin object string, extrapolates from the construction node address along the building construction edge to the observation component node, writes the on-site state of the observation component node, the recent intervention bit and the construction layer transmission bit into the state base segment by segment, performs binary expansion on the unobserved state bits, and outputs the candidate hidden state string. The quantifier elimination module is used to write the candidate hidden state into the free integer position through the candidate hidden state string, and to connect the action position through the action cell and the entry cell to form the existence quantifier position. The Cooper Pressburg arithmetic quantifier elimination is performed by the connector, the order, and the remainder, and the output quantifier-free action signature is output. The path matching module is used to generate action difference words by bitwise XORing the action signature without quantifiers. When all bits of the action difference word are zero, no evidence collection management record is output. When the action difference word contains non-zero bits, an evidence collection bipartite graph is constructed with the entry bit and the non-zero bits, and Hungarian bipartite graph matching is performed with the entry cost word to output the evidence collection path record. The decision write-back module is used to generate evidence collection result records based on the evidence collection path records, perform position matching between the evidence collection result records and candidate hidden state strings, and write back the twin state of the target hidden part and the maintenance or repair management record of the building component product when a match is found, and write back the review twin state when a match is not found, and output the management decision record before repair.

[0014] The technical effects and advantages of this invention are as follows: 1. This solution first determines the necessity of evidence collection based on the action difference words, and then generates the evidence collection path, so that hidden parts are no longer directly supplemented by data missing, which relatively reduces meaningless inspection, opening and expansion of the sealing area; 2. The hidden part number is converted into a construction node address through the building part code, so that the on-site operation record falls into the digital twin construction node, which improves the location of hidden parts and the subsequent status handover. 3. Perform Cooper Pressburg arithmetic quantifier elimination on the candidate hidden state string and generate quantifierless action signatures, so that the unobservable state is transformed into a basis for judging action differences; 4. Construct a bipartite graph for evidence collection between the entry point and the non-zero position of the action difference, and perform Hungarian matching with the entry cost word to make the evidence collection path generation take into account the effects of component passage, surface contact and spatial closure. 5. Based on the matching of the status word of the evidence collection result with the candidate hidden status, and write back the twin status and maintenance records, so that the management decision before the repair has the status source and record inheritance. Attached Figure Description

[0015] Figure 1 This is a flowchart outlining the method steps of the present invention; Figure 2 This is a schematic diagram of the system module structure of the present invention. Detailed Implementation

[0016] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0017] Refer to the instruction manual appendix Figure 1-2 The present invention provides a dynamic management method for the entire life cycle of ancient buildings based on digital and intelligent technologies, comprising: S1. Obtain the digital twin base map of the ancient building and the on-site work record before the renovation. Convert the hidden part number into the construction node address using the building part code, and convert the handling actions in the work task into handling action bits. Output the hidden part twin object string. In this embodiment, S1 is used to input the hidden part information in the pre-renovation site work record into the construction node with a definite address in the digital twin base map of the ancient building, and to convert the disposal actions in the renovation task into disposal action bits that participate in subsequent state derivation and action signature generation; during processing, the digital twin base map of the ancient building provides the building hierarchy tree and component assembly relationship, and the pre-renovation site work record provides the hidden part number and operation action name. Then, through hierarchical bit width splicing, segmented bit matching, and action sequence access, a hidden part twin object string is formed, which is used by S2 to read the construction node address of the target hidden part; this implementation process includes the following steps: In S11, the building component coding table is used to generate addresses for reading in segments according to the hierarchy for the structural nodes in the digital twin base map of ancient buildings; the input is the building hierarchy tree in the digital twin base map of ancient buildings. The building hierarchy tree is generated by the building space hierarchy and component assembly hierarchy in the digital twin base map of ancient buildings. The root node is written into the ancient building unit, and the child nodes are expanded step by step according to the courtyard area, building unit, roof area, wall area, beam frame area, component layer and hidden part layer. During processing, first read all child nodes under the same parent node, and write each child node into the current level according to the edge writing order from the parent node to the child node. Then calculate the bit width of the current level based on the total number of child nodes under the same parent node. When the total number of child nodes is one, write one bit into the current level bit width. When the total number of child nodes is greater than one, the current level bit width takes the number of binary bits required to represent the total number of child nodes minus one. Then, the parent node address is shifted left by the current level bit width and then connected to the current level sequence to form the child node address. The parent node address, current level sequence, current level bit width, and child node address are written into the building part coding table. When there are component numbers in the digital twin base map of ancient buildings, the component numbers are written into the node initial field and do not replace the child node address generated by the parent node address and current level sequence. In S12, the node address is constructed to place the hidden part number in the pre-renovation site work record into the end node of the digital twin base map of the ancient building; the input is the pre-renovation site work record and the building part coding table; During processing, when recording the hidden part number in the on-site work record before repair, the hidden part number is segmented and the position is taken from high to low according to the level bit width in the building part coding table. Each segmented position value is matched sequentially with the child node of the same level in the building level tree. When all levels hit a single child node, the address of the last child node is read as the construction node address. When recording the names of artificial parts in the on-site work record before the renovation, first convert the names of artificial parts into the corresponding hierarchical order through the building part name table, and then splice them together according to the hierarchical position width in the building part coding table to form a hidden part number before performing segmented position matching. When there is a no hit or multiple hits, the corresponding hidden part number is not written into the hidden part twin object string. Instead, the hidden part number, segment position value, number of hit nodes and on-site operation record order are written into the placement verification record for manual review or reading when the on-site operation record is corrected before the next round of repair. In S13, the action bit is used to convert the operation action name into a bit string object for subsequent state expansion reading; the input is the on-site operation record before repair and the ancient building repair operation sequence table that have hit the construction node address; During processing, the names of the work actions are read from the work tasks bound to the on-site work records before the renovation. Then, the order of the disposal actions is generated based on the row order of the work action names in the ancient building renovation work sequence table. The ancient building renovation work sequence table comes from the work plan table before the renovation or the protection and management action registration table. When the same work action name appears repeatedly, the row order of the first registration is read. The new work action name is written into the action verification record and the generation of disposal action positions is paused. When the action sequence is valid, read the width of the construction node address, shift the action sequence to the left by the width of the construction node address and then connect it to the construction node address to form the action bit. Then, connect the action bit to the end of the construction node address and output the hidden part twin object string. Write the hidden part twin object string into the construction node address, the action bit and the on-site operation record sequence, so that S2 can read the target hidden part along the edge of the building structure. Through the aforementioned processing, S1 converts the relationship between the digital twin base map of the ancient building, the on-site work records before the renovation, and the work tasks into a hidden part twin object string with structural node addresses and disposal action bits, thus solving the problems of different sources of hidden part names, component numbers not being able to directly participate in bit string calculation, and repeated registration of work actions leading to the inability to determine disposal actions. In practical applications: Taking the beam end wall-entry part of a wooden ancient building as an example, the digital twin base map of the ancient building first generates a building hierarchy tree by generating the building unit, beam frame area, beam components and beam end hidden parts level by level. The "beam end wall-entry point of the east side main room" in the on-site operation record before the repair is converted into a hidden part number by the building part name table, and then the structural node address of the beam end hidden part is hit by segmenting according to the hierarchical position width. The "partial inspection" in the operation task reads the row order from the ancient building repair operation sequence table to generate the disposal action order, and splices it with the structural node address to form the disposal action position, and finally writes it into the hidden part twin object string, so that the subsequent S2 can continue to deduce the candidate hidden state from the structural node address of the beam end hidden part.

[0018] S2. Based on the target hidden part in the twin object string of hidden parts, push the construction node address outward along the building construction edge to the observation component node, write the on-site state of the observation component node, the recent intervention bit and the construction layer transmission bit into the state base segment by segment, perform binary expansion on the unobserved state bits, and output the candidate hidden state string. In this embodiment, S2 is used to obtain the construction node address of the target hidden part from the hidden part twin object string, and push it outward from the building construction edge in the ancient building digital twin base map to the observation end to obtain the observation component node that has a construction connection relationship with the target hidden part. Then, the on-site state of the observation component node, the recent intervention bit of the target hidden part, and the construction layer transmission bit are written into the state base. Finally, the unobserved state bits in the state base are binary expanded to form a candidate hidden state string, which is used by S3 to generate a wordless action signature. This implementation process includes the following steps: In S21, the extrapolation path is used to locate the observation component node with the on-site state source from the target hidden part; the input is the hidden part twin object string and the ancient building digital twin base map; During processing, the construction node address of the target hidden part is extracted from the twin object string of the hidden part, and the construction node address is used as the current node address. The building construction edge with the same starting address and current node address and the direction pointing to the observation end is read in the digital twin base map of the ancient building. The direction is generated by the construction connection relationship from the hidden layer to the visible layer. After reading the building construction edge, write the endpoint address of the building construction edge into the extrapolation path bit and write the current node address into the extrapolation read address table; if the endpoint address does not have the field status bit, first determine whether the endpoint address already exists in the extrapolation read address table. If the endpoint address already exists in the extrapolation read address table, stop reading the current building construction edge and write it into the extrapolation loop record. If the endpoint address does not exist in the extrapolation read address table, replace the current node address with the endpoint address and continue reading. When the endpoint address has a field status bit, stop extrapolation and output the observation component node address; when there are multiple direction bits pointing to the building structure edge of the observation end at the same current node address, read them one by one according to the writing order of the building structure edge, output the endpoint address with the field status bit first, and when none of the endpoint addresses have the field status bit, continue extrapolation according to the writing order to access the extrapolation path bit; when there is no unread building structure edge and no observation component node address is obtained, write the target hidden part into the observation missing item record and stop generating the state base; In S22, the observation acceptance record is used to connect the on-site status of the observed component node and the recent intervention information of the hidden part of the target to the same status source; the input quantities are the address of the observed component node, the on-site operation record before repair, and the repair sequence table; During processing, the field status bit is read based on the node address of the observed component. The field status bit is generated by encoding the visible defects mark, component integrity mark, surface dampness mark, crack location mark or manual verification mark in the field operation record before repair. When there are multiple field operation records for the same node address of the observed component, the most recent operation item is read in reverse order of operation time. When the operation times are the same, the next record in the order of the field operation record is read. Then, in the repair sequence table, read the work items that the construction node address hit in reverse order of the work time. When a hit work item exists, shift the work order of the hit work item to the left by the work type bit width and then connect it to the work type bit to generate the most recent intervention bit. When multiple work items are hit, read the work item after the work time. When the work times are the same, read the work item after the work order. When no work item matches the structural node address in the repair sequence table, an empty intervention code is written to the nearest intervention bit; the field status bit and the nearest intervention bit are written to the observation acceptance record for reading when the structural layer transfer bit is generated. In S23, the construction layer transfer position is used to express the hierarchical transfer direction between the target hidden part and the observation component node, and together with the field status position and the near intervention position, it forms the state basis; the input quantities are the construction layer number of the target hidden part, the construction layer number of the observation component node, and the observation acceptance record. During processing, the structural layer number is generated from the layer where the structural node is located in the digital twin base map of the ancient building. The hidden part layer, base layer, load-bearing layer, surface layer and visible layer are written into the structural layer number in order from the inside to the outside. After reading the structural layer number of the target hidden part and the structural layer number of the observed component node, the structural layer difference value is generated by subtracting the structural layer number of the observed component node from the structural layer number of the target hidden part. When the structural layer difference value is negative, the sign bit is written as 1, and when the structural layer difference value is zero or positive, the sign bit is written as zero. Then, the sign bit of the structural layer difference value is connected to the absolute value bit of the structural layer difference value to the left to generate the structural layer transmission bit. Then, the state base is written in the order of bit width of the on-site state bit, the recent intervention bit, and the structural layer transmission bit, and the bit segment occupied by the state of the target hidden part itself is written into the empty bit as the unobserved state bit; if the structural layer number of the target hidden part or the structural layer number of the observation component node is empty, the target hidden part is written into the hierarchical verification record and is not entered into the candidate hidden state string generation. In S24, the candidate hidden state string is used to list all the bit value branches of the target hidden part before the evidence is collected, and to provide candidate input for the generation of free integer bits in S3; the input is the state basis; during processing, the number of unobserved state bits with empty values ​​in the state basis is counted first and the unobserved count is generated. The field state bits, recent intervention bits and construction layer transmission bits are all treated as generated bits and do not participate in binary expansion. Then, starting from zero, binary expansion numbers are generated incrementally by 1 until the binary expansion number reaches 2 raised to the power of the unobserved count minus 1. If each binary expansion number is less than the width of the unobserved count, zeros are added to the left and the unobserved state bits in the state base are filled in bit by bit in the order of the unobserved state bits from the high bit to the low bit. Each binary expansion number generates a candidate hidden state. When the number of unobserved counts is zero, the state basis is directly written into the candidate hidden state string; after the candidate hidden state is generated, deduplication is performed according to the candidate number. For duplicate candidate hidden states, the candidate number written first is retained, and the duplicate candidate number is written into the candidate deduplication record. The candidate hidden state string is then output. Through the aforementioned processing, S2 transforms the construction node address in the hidden part twin object string into a candidate hidden state string composed of the expansion results of the observed component node field state, recent intervention bit, construction layer transmission bit and unobserved state bit, so that S3 can generate a corresponding wordless action signature for each candidate hidden state. In practical applications: Taking the beam end of a wooden ancient building entering the wall as an example, the structural node address of the target hidden part is first pushed outward along the internal structural edge of the wall towards the wall observation end. After reading the wall crack observation node with the on-site status bit, the outward push stops. The system reads the on-site status bit of the wall crack from the pre-repair on-site operation record and generates the most recent intervention bit from the most recent "partial inspection" operation item of the beam end entering the wall from the repair sequence table. Then, the structural layer transfer bit is generated by subtracting the structural layer number of the wall observation node from the structural layer number of the beam end entering the wall. The unobserved status bit of the beam end entering the wall itself is expanded in binary to form a candidate hidden status string, which is used to judge whether different hidden states will change the pre-repair management actions.

[0019] S3. Write the candidate hidden state into the free integer bits through the candidate hidden state string, and connect the action bit through the action grid and the entry grid to form the existence quantifier bit. Then, the Cooper Pressburg arithmetic quantifier elimination is performed by the connector, order, and remainder items to output the quantifier-free action signature. In this embodiment, S3 is used to convert each candidate hidden state in the candidate hidden state string into a quantifierless action signature that can be used for action difference comparison. This allows the changes in the pre-repair management actions to be deduced from the building structure edge, the action cell, the entry cell, and the disposal action cell even before direct evidence is obtained. During processing, the candidate hidden states are first substituted into the pre-repair disposal action table to generate existence quantifier cells. Then, the structural connection relationship from the target hidden part to the action cell, the sequential relationship between the entry cell and the action cell, and the remainder relationship between the disposal action cell and the building cell modulus are written into the Pressburg constraint string. Subsequently, the intermediate unobservable variables are deleted by Cooper Pressburg arithmetic quantifier elimination, forming a quantifierless action signature for S4 to perform bitwise XOR. This implementation process includes the following steps: In S31, the free integer bits and the existence quantifier bits are used to establish a computable integer relationship between the candidate hidden state and the pre-repair management action; the input quantities are the candidate hidden state string, the construction node address of the target hidden part, and the pre-repair handling action table. The pre-repair handling action table is generated from the handling action order, the address of the building part to be affected, and the operation entry position in the pre-repair work plan table. During processing, the candidate hidden status is read in ascending order of candidate number. The candidate number is written to the first part of the free integer position, and the candidate hidden status is written to the second part of the free integer position. Then, the status position in the candidate hidden status is matched line by line with the status position in the pre-repair treatment action table. When a line is matched, the order of the same-line treatment action is read to generate the treatment action position, the address of the building part of the same-line action is read to generate the action position, and the entry position of the same-line operation is read to generate the entry position. The building grid width is taken from the longest bit width of the building part code in the digital twin base map of the ancient building, and the entry grid width is taken from the longest bit width of the entry spatial node address. When the active grid and the entry grid are insufficient for the corresponding bit width, zeros are added to the left. Then, the action position is shifted left by the width of the building cell and connected to the action cell, then shifted left by the width of the entry cell and connected to the entry cell, and the existence quantifier position is output; when the candidate hidden state does not hit the action table before repair, the candidate number and the candidate hidden state are written into the action missing record and are not included in the Presburg constraint string generation. In S32, the Pressburg constraint string is used to transform building structure reachability relations, repair operation sequence relations, and disposal action coding relations into constraint objects containing only integer addition, order, and remainder; the input quantities are existence quantifiers, target hidden part construction node addresses, action cells, and entry cells; During processing, the address of the node constructed at the hidden part of the target is used as the starting address. The construction edge of the building is read along the digital twin base map of the ancient building to the active grid. The edge number of each construction edge in the path is accumulated according to the reading order to form a connection item. When multiple paths exist, connection item branches are generated separately. The order term is generated by subtracting the action term from the entry term. The order term retains the sign and absolute value. A negative value indicates that the entry term is before the action term, a zero value indicates that the entry term and the action term are in the same position, and a positive value indicates that the entry term is after the action term. The building term modulus is generated by raising the building term width to the power of two. The remainder term is generated by taking the modulus of the action term with respect to the building term modulus. Then, the connection branch, order term, and remainder term are written into the Pressburg constraint string according to the candidate number; if no building construction edge leading to the action cell is read, the candidate number is written into the construction unreachable record and is not entered into the Cooper Pressburg arithmetic quantifier elimination; In S33, Cooper Pressburg arithmetic quantifier elimination is used to replace existential quantifiers in the Pressburg constraint string with quantifier-less terms represented by free integers and remainder branches; the input is the Pressburg constraint string. During processing, first read the linear terms containing existential quantifiers. If the coefficient of the existential quantifier is zero, retain the linear term and read the next linear term. If the coefficient of the existential quantifier is not zero, multiply the linear terms containing existential quantifiers in the same Pressburg constraint string by the product of the absolute values ​​of the coefficients of the existential quantifiers, so that the existential quantifiers move to the same side of the inequality and form a unified term. Then, expand the divisibility relation in the remainder terms into remainder branches. Write the remainder branches starting from zero and incrementing by one until the remainder branches reach the building cell modulus minus one. Then, the lower bound of the existential quantifier position is generated by the connector branch, and the upper bound of the existential quantifier position is generated by the order term. When the lower bound is greater than the upper bound, the corresponding remainder branch is written into the non-action branch. When the lower bound is not greater than the upper bound, the existential quantifier position is replaced by the lower bound plus remainder branch, and the replaced non-quantifier branch is output. The non-action branch does not enter the generation of non-quantifier action signature. In S34, quantifier removal and branch recursion are used to ensure that all intermediate quantifier positions in the Pressburg constraint string are removed before entering action signature compression; the input is the quantifier-free branch and the constraint branch that still contains quantifier positions output by S33. During processing, after replacing the existing quantifier in each remainder branch, the existing quantifier is deleted, and the number of remaining existing quantifiers in the current Pressburg constraint string is counted. When the number of remaining quantifier positions is zero, a quantifier-free constraint string is formed by concatenating the candidate sequence number, action position, action position, entry position, connector, order, and remainder position in a fixed order; when the number of remaining quantifier positions is not zero, the next quantifier position is read and S33 is returned to perform normalization, remainder expansion, boundary replacement, and quantifier deletion. When all candidate hidden states are written as unreachable branches or unreachable records are constructed, the target hidden part is written as the constraint conflict management record, and the output of unquantifier action signatures to S4 is stopped. In S35, the back-substitute action bit is used to check whether the unqualified constraint string maintains the bit value relationship of the action bit, and compresses the unqualified constraint string that meets the bit value relationship into an unqualified action signature; the input is the unqualified constraint string and the candidate hidden state string; During processing, the candidate hidden state corresponding to the same candidate number is replaced by the unqualified constraint string. The replacement action bit is generated according to the disposal action bit segment in the unqualified constraint string. Then, the replacement action bit and the disposal action bit generated by S31 are XORed bit by bit. When the XOR result is all zeros, the candidate sequence number, action bit, action cell, entry cell, and remainder item are concatenated and written into the quantifier-free action signature; when the XOR result contains non-zero bits, the candidate sequence number, back substitution action bit, action bit, and XOR result are written into the conflict elimination record, and the candidate hidden state is not included in the quantifier-free action signature set; when the quantifier-free action signature set is empty, the target hidden part is written into the review twin state and the conflict elimination record is retained. Through the aforementioned processing, S3 converts the unobservable state branches in the candidate hidden state string into wordless action signatures, enabling S4 to directly compare whether different candidate hidden states lead to differences in the pre-repair management actions. In practical applications: Taking the beam end wall entry location as an example, the candidate hidden states "beam end damp" and "beam end decayed" respectively match different operation actions in the pre-repair treatment action table. The system concatenates the corresponding treatment action position, the wall internal action grid, and the inspection entrance grid into an existence quantifier position; then, it accumulates the edge position number along the building structure edge from the beam end wall entry location to the wall action grid to generate a connection item, and generates an order item based on the difference between the inspection entrance grid and the action grid, and generates a remainder item based on the modulus of the treatment action position divided by the building grid; after the Cooper Pressburg arithmetic quantifier eliminates and deletes the existence quantifier position, it outputs a quantifier-free action signature for S4 to determine whether different hidden states need to enter the evidence collection path matching.

[0020] S4. Generate action difference words by bitwise XORing the action signature without quantifiers. When all bits of the action difference word are zero, output no evidence collection management record. When the action difference word contains non-zero bits, construct an evidence collection bipartite graph with the entry bit and non-zero bits, and perform Hungarian bipartite graph matching with the entry cost word, and output evidence collection path record. In this embodiment, S4 is used to determine whether different candidate hidden states cause differences in pre-repair management actions based on the action signature without quantifiers. When differences exist, the difference location, entrance location, and building construction path are organized into an evidence collection bipartite graph, and an evidence collection path record is generated through Hungarian bipartite graph matching. During processing, the action signature without quantifiers is first XORed bitwise to form an action difference word. The non-zero bits in the action difference word are then mapped to the action grid. Subsequently, the left vertex is generated using the entrance location leading to the target hidden part, the right vertex is generated using the non-zero bits in the action difference word, and the entrance cost word is formed using the components passed through, contact surfaces, uncovering, and enclosed spaces in the building construction path for reading by Hungarian bipartite graph matching. This implementation process includes the following steps: In S41, the action difference word is used to express whether there is a difference in the pre-repair management action between the candidate hidden states; the input is the set of action signatures without quantifiers output by S3; during processing, the bit width of each action signature without quantifiers is read first, and the last bit of the bit width value is taken as the standard bit width of the signature. Short action signatures without quantifiers are padded with zeros on the left before entering the comparison. Then, the action signatures without quantifiers are read in pairs according to the order of the candidate hidden states to form action signature pairs. For each action signature pair, perform a bitwise XOR operation to generate candidate difference words, and then perform a bitwise OR operation on all candidate difference words to obtain the action difference word. When all bits of the action difference word are zero, it means that the action bits, action grid bits, and entry grid bits corresponding to different candidate hidden states have no bit value differences. Output no evidence collection management record and write the target hidden part into the twin state to be observed. When the action difference word contains non-zero bits, the action difference word is written into the difference trigger record for S42 to generate the right vertex of the evidence collection bipartite graph. In S42, the evidence-gathering bipartite graph is used to establish candidate correspondences between the non-zero bits in the action difference word and the entrance positions that can reach the hidden part of the target; the inputs are the action difference word, the action signature without quantifiers, and the digital twin base map of the ancient building. During processing, first read the non-zero position in the action difference word according to the fixed boundaries of the disposal action position segment, the action position segment, and the entry position segment in the action signature without quantifier. When the non-zero position falls into the action position segment, directly read the corresponding action position of that segment. When the non-zero position falls into the disposal action position segment or the entry position segment, read the action position in the same action signature without quantifier as the non-zero position associated action position. Then, the access point node leading to the hidden part of the target is read from the digital twin base map of the ancient building. The access point node address is written into the access point and the left vertex is generated. The left vertex is not generated when the access point node does not have the access permission bit. Each non-zero bit in the action difference word is written into the right vertex. The building structure path is formed by reading the building structure edge from the entrance to the target hidden part. When the node address set of the building structure path contains non-zero associated cells, it is written into the bipartite graph edge. When the node address set does not contain non-zero associated cells, it is not written into the bipartite graph edge. When there are multiple building structure paths from the same entrance to the target hidden part, each building structure path generates a bipartite graph edge. In S43, the entry cost word is used to convert the occupancy of ancient building components and open spaces by the evidence collection path into the matching cost value of Hungarian bipartite graph matching; the input is the bipartite graph edge and the building construction path corresponding to the bipartite graph edge; During processing, the system reads the target hidden part from the entrance along the building structure path. The number of component nodes in the path is written into the component count, the number of surface nodes with contact permission positions in the path is written into the contact surface count, the number of covering layer nodes in the path is written into the uncover count, and the number of space nodes that need to be temporarily closed in the path is written into the closed space count. After the component count data is written to the front of the entry cost word, the contact surface count data is shifted to the left by the component count data bit width and then entered into the entry cost word. The uncovering count data is shifted to the left by the contact surface count data bit width and then entered into the entry cost word. The enclosed space count data is shifted to the left by the uncovering count data bit width and then entered into the entry cost word. The entry cost word is read as an unsigned integer as the matching cost value. When the building construction path lacks component node, surface node, cover layer node, or space node fields, the bipartite graph edge is written into the entry verification record and does not enter the matching cost value generation. In S44, the label initialization is used to transfer the entry cost word into the relaxation calculation in the Hungarian bipartite graph matching; the inputs are the left vertex, right vertex, bipartite graph edge, and matching cost value; during processing, for each left vertex, the matching cost value on the bipartite graph edge associated with that left vertex is read, and the previous matching cost value is read in ascending order as the left vertex label; the right vertex label of each right vertex is written to zero. Then, a relaxation word is calculated for each bipartite edge. The relaxation word is obtained by subtracting the left and right vertex labels of the two ends of the same bipartite edge from the matching value. When the relaxation word is zero, the bipartite edge is written into the zero-relaxation edge set. When the relaxation word is less than zero, the corresponding bipartite edge is written into the label exception record and removed from the zero-relaxation edge set of the current round. When there are multiple matching values ​​for the same left vertex, the previous bipartite edge is read according to the bipartite edge generation order. In S45, alternating tree expansion and match flipping are used to generate matching relationships from entry positions to non-zero action difference positions from the zero-relaxation edge set; the inputs are the zero-relaxation edge set and the current matching table; during processing, the unmatched left vertex is selected as the root node of the alternating tree, and the right vertex connected to the left vertex in the alternating tree is read along the zero-relaxation edge set. When an unmatched right vertex is read, backtrack along the alternating tree path from the unmatched right vertex to the root node of the alternating tree, write a match for the unmatched bipartite edges on the path, and delete the match for the matched bipartite edges on the path to complete the match reversal; When a matched right vertex is read, the left vertex connected to the matched right vertex in the current matching table is read, and the connected left vertex is added to the alternation tree before continuing to read the zero-relaxation edge set; the left and right vertices already read in the same round of the alternation tree are written to the alternation tree read table, and will not be added again in subsequent expansions; In S46, label rewriting is used to generate new zero-relaxation edges when the alternating tree cannot reach the unmatched right vertex, and outputs the evidence path record after the action difference non-zero bit is overwritten; the inputs are the alternating tree, the set of zero-relaxation edges, the relaxation word, and the current matching table; During processing, when no unmatched right vertex is read in the alternating tree, the relaxation words of the bipartite graph edges between the left vertex inside the alternating tree and the right vertex outside the alternating tree are read. The previous relaxation word is read in ascending order of value as the label rewriting word. The label rewriting word is added to the label of the left vertex inside the alternating tree, and the label rewriting word is subtracted from the label of the right vertex inside the alternating tree. Then, return to S44 to regenerate the zero relaxation edge set. If there is no right vertex outside the alternating tree or no relaxed word that can generate a label rewriting word, and the number of matched right vertices in the current matching table does not reach the number of non-zero bits in the action difference word, then output the uncovered difference record and write the non-zero bits corresponding to the unmatched right vertex into the verification twin state. When the number of matched right vertices reaches the number of non-zero bits in the action difference word, read the entry position, building construction path, entry cost word and matching cost value in the current matching table according to the right vertex order, and output the evidence collection path record. Through the aforementioned processing, S4 converts the action differences between action signatures without quantifiers into an evidence-gathering bipartite graph, and generates evidence-gathering path records by matching the entry cost word with the Hungarian bipartite graph, so that the target hidden part only enters the evidence-gathering path generation when the candidate hidden state causes the difference in management actions before repair. In practical applications: Taking the beam end wall entry location as an example, the candidate hidden states "beam end damp" and "beam end decayed" generate non-zero bits in the action signature of the action grid segment. The system generates entrance positions for the wall inspection port, beam frame access node and temporary work station respectively, and determines whether the building construction path from each entrance position to the beam end wall entry location covers the non-zero bit associated action grid. Then, the number of beam frame components passed through, the number of wall surfaces contacted, the number of surface layers removed and the number of closed room grids are concatenated into the entrance cost word. The corresponding evidence collection path record is output through Hungarian bipartite graph matching, which is used by S5 to form the evidence collection result record.

[0021] S5. Based on the evidence collection path record, form the evidence collection result record, perform position matching between the evidence collection result record and the candidate hidden state string, and write back the twin state of the target hidden part and the maintenance or repair management record of the building component product when the match is found, and write the review twin state when the match is not found, and output the pre-repair management decision record. In this embodiment, S5 is used to convert the on-site backfill information generated after the evidence collection path record is executed into the evidence collection result record of the target hidden part, and to complete the twin state rewriting and writing of the building component product maintenance or repair management record by comparing the bit values ​​of the evidence collection result status word and the candidate hidden status string; during processing, the evidence collection path record provides the address and path order of the target hidden part, the evidence collection operation backfill item provides the original observation code, and the candidate hidden status string provides the candidate hidden status bit width and candidate sequence number. Then, after bit width padding, same bit XOR, hit writing and miss verification, the pre-repair management decision record is output; this implementation process includes the following steps: In S51, the evidence collection result status word is used to convert the on-site evidence collection result into a bit string that can be compared with the candidate hidden status string; the input is the evidence collection path record and the evidence collection operation backfill item. The evidence collection operation backfill item is generated by the on-site review record after the evidence collection path record is executed. The on-site review record is written with the path sequence, target hidden part address, observation part address, original observation code and backfill time. During processing, first read the target hidden part address and path order from the evidence collection path record, then retrieve the evidence collection operation backfill item with the same path order, and then read the candidate hidden state bit width of the candidate hidden state string. When the width of the original observation code is greater than the width of the candidate hidden state, the lower right bits of the original observation code are retained and truncated to the width of the candidate hidden state. When the width of the original observation code is less than the width of the candidate hidden state, zeros are padded to the left of the original observation code to the width of the candidate hidden state to generate the evidence collection result status word. The status word of the evidence collection result is appended to the end of the target hidden part address, and the evidence collection result record is output; when the original observation code is not written into the evidence collection operation backfill item, the path sequence and the target hidden part address are written into the backfill verification record, and no evidence collection result status word is generated. In S52, the hit candidate sequence number is used to determine the candidate hidden state corresponding to the evidence result status word, so that the twin state rewriting of the target hidden part has a candidate source; the input is the evidence result record and the candidate hidden state string; During processing, first read the evidence result status word in the evidence result record, then read the candidate hidden status in the candidate hidden status string by incrementing the candidate sequence number; for each candidate hidden status read, perform a bitwise XOR operation between the evidence result status word and the candidate hidden status, and perform a full bitwise OR operation on the bitwise XOR result. When the full bitwise OR result is zero, write the hit candidate sequence number and stop reading. If all bits or the result is not zero, continue reading the next candidate hidden state; when the candidate hidden state string is generated, it is deduplicated according to the candidate sequence number. For duplicate candidate hidden states, the candidate sequence number written first is retained, and the duplicate candidate sequence number is written to the candidate deduplication record; when the candidate hidden state reading ends and no matching candidate sequence number is written, a miss flag is output, and the evidence collection result record is retained for reading when expanding the candidate hidden state string in the next round. In S53, the pre-renovation management decision record is used to write the hit or miss results into the digital twin base map of the ancient building, and simultaneously form the maintenance or repair management record of the building component products; the input quantities are the hit candidate sequence number, miss mark, evidence collection path record, evidence collection result record and the digital twin base map of the ancient building; During processing, when a candidate number is present, the hidden state of the candidate corresponding to the candidate number is read, and the twin state slot in the digital twin base map of the ancient building is located by the address of the target hidden part. The state bit to be verified in the twin state slot is rewritten to the hidden state of the candidate. Then, the target hidden location address, evidence collection path record, evidence collection result status word, hit candidate sequence number, twin status before rewriting value and twin status after rewriting value are entered into the building component product maintenance or repair management record according to the path order. When a miss flag exists, the pending status bit in the twin status slot is rewritten to the verification twin status, and the evidence collection result record is appended to the end of the verification twin status; finally, the target hidden part address, evidence collection path record, hit candidate sequence number or miss flag, and twin status slot are written into the result and spliced ​​into the pre-repair management decision record. Through the aforementioned processing, S5 transforms the on-site backfill information after the execution of the evidence collection path record into the twin state rewriting result of the target hidden part, and writes the hit result or the non-hit result into the building component product maintenance or repair management record, so that the pre-repair management decision record can take over the evidence collection path record output by S4. In practical applications: Taking the beam end wall-entry part as an example, the evidence collection path record specifies that the evidence is collected from the wall inspection port and the hidden part of the beam end is collected. The on-site verification record fills in the original observation code according to the path sequence. The system truncates or pads the original observation code with zeros according to the candidate hidden state bit width to form the evidence collection result status word, and then compares it bit by bit with the candidate hidden states corresponding to "beam end damp" and "beam end decay". If the evidence collection result status word matches the candidate sequence number of "beam end decay", the state bit to be verified of the beam end wall-entry part in the ancient building digital twin base map is rewritten to the state of beam end decay, and the evidence collection path record, evidence collection result status word and the matched candidate sequence number are written into the building component product maintenance or repair management record.

[0022] Furthermore, the present invention also includes a dynamic management system for the entire life cycle of ancient buildings based on digital and intelligent means. The system includes a twin positioning module, a state expansion module, a quantifier elimination module, a path matching module, and a decision writing-back module. The twin positioning module is used to obtain the digital twin base map of the ancient building and the on-site work record before the renovation. It converts the hidden part number into the construction node address by the building part code, and converts the disposal actions in the work task into disposal action bits, and outputs the hidden part twin object string. The state expansion module, based on the target hidden part in the hidden part twin object string, extrapolates from the construction node address along the building construction edge to the observation component node, writes the on-site state of the observation component node, the recent intervention bit and the construction layer transmission bit into the state base segment by segment, performs binary expansion on the unobserved state bits, and outputs the candidate hidden state string. The quantifier elimination module is used to write the candidate hidden state into the free integer position through the candidate hidden state string, and to connect the action position through the action cell and the entry cell to form the existence quantifier position. The Cooper Pressburg arithmetic quantifier elimination is performed by the connector, the order, and the remainder, and the output quantifier-free action signature is output. The path matching module is used to generate action difference words by bitwise XORing the action signature without quantifiers. When all bits of the action difference word are zero, no evidence collection management record is output. When the action difference word contains non-zero bits, an evidence collection bipartite graph is constructed with the entry bit and the non-zero bits, and Hungarian bipartite graph matching is performed with the entry cost word to output the evidence collection path record. The decision write-back module is used to generate evidence collection result records based on the evidence collection path records, perform position matching between the evidence collection result records and candidate hidden state strings, and write back the twin state of the target hidden part and the maintenance or repair management record of the building component product when a match is found, and write back the review twin state when a match is not found, and output the management decision record before repair.

[0023] Working principle: This scheme first converts the building hierarchy and component relationships in the digital twin base map of the ancient building into calculable structural node addresses. Then, it assigns the hidden parts in the pre-renovation on-site work records to the corresponding structural nodes and converts the disposal actions in the work tasks into disposal action positions. Subsequently, it traces outward from the target hidden part along the building structure edge to the observable component nodes, reads the on-site status, recent interventions, and structural layer transmission relationships, and expands to form candidate hidden states. Different candidate hidden states then enter the Cooper Pressburg arithmetic quantifier elimination process to be converted into quantifierless action signatures, which are used to determine whether different hidden states will change the pre-renovation management actions. If there is no difference in the action signatures, no evidence collection is arranged. If there is a difference, an evidence collection bipartite graph is constructed, and an evidence collection path is generated through Hungarian bipartite graph matching. Finally, the evidence collection results are written back to the ancient building digital twin status and the building component maintenance management records. In practical applications, for example, when damp marks appear on the wall at the beam end of a wooden ancient building, the solution does not directly demolish or inspect the wall to complete the data. Instead, it first locates the beam end at the wall in the digital twin base map, and then generates candidate states such as no abnormalities at the beam end, damp beam end, and decayed beam end based on the wall observation status, historical repair records, and structural layer relationships. The system further calculates whether these candidate states will lead to changes in management actions such as continued observation, partial inspection, or support repair. If the management actions corresponding to different states are the same, it is recorded as no evidence collection management. If different states will change the handling results, the evidence collection path is calculated from entrances such as wall inspection openings and beam frame passage positions. The path with controlled impact on component contact, covering layer removal, and space closure is selected for evidence collection, and the finally confirmed beam end state is written back to the digital twin base map and maintenance records.

[0024] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A dynamic management method for the entire life cycle of ancient buildings based on digital and intelligent technologies, characterized in that: include: S1. Obtain the digital twin base map of the ancient building and the on-site work record before the renovation. Convert the hidden part number into the construction node address using the building part code, and convert the handling actions in the work task into handling action bits. Output the hidden part twin object string. S2. Based on the target hidden part in the twin object string of hidden parts, push the construction node address outward along the building construction edge to the observation component node, write the on-site state of the observation component node, the recent intervention bit and the construction layer transmission bit into the state base segment by segment, perform binary expansion on the unobserved state bits, and output the candidate hidden state string. S3. Write the candidate hidden state into the free integer bits through the candidate hidden state string, and connect the action bit through the action grid and the entry grid to form the existence quantifier bit. Then, the Cooper Pressburg arithmetic quantifier elimination is performed by the connector, order, and remainder items to output the quantifier-free action signature. S4. Generate action difference words by bitwise XORing the action signature without quantifiers. When all bits of the action difference word are zero, output no evidence collection management record. When the action difference word contains non-zero bits, construct an evidence collection bipartite graph with the entry bit and non-zero bits, and perform Hungarian bipartite graph matching with the entry cost word, and output evidence collection path record. S5. Based on the evidence collection path record, form the evidence collection result record, perform position matching between the evidence collection result record and the candidate hidden state string, and if a match is found, write back the twin state of the target hidden part and the maintenance or repair management record of the building component product; if a match is not found, write the review twin state and output the pre-repair management decision record.

2. The method for dynamic management of the entire life cycle of ancient buildings based on digital and intelligent means according to claim 1, characterized in that: S1 includes: S11. Read the building hierarchy tree from the digital twin base map of ancient buildings, write the child node into the current level sequence according to the edge order from the parent node to the child node, use the binary code length of the number of nodes in the current level as the bit width of the current level, shift the address of the parent node to the left of the current level bit width and then connect it to the current level sequence, and output the building part coding table. S12. Read the hidden part number from the site work record before the repair, perform high-level segmentation on the hidden part number according to the level bit width in the building part coding table, and match the child node order in the building hierarchy tree with the segmented position value in sequence. When a single child node is hit, read the child node address as the construction node address. When there is no hit or multiple hits, write the placement verification record. S13. Read the operation action name from the pre-renovation site operation record, generate the disposal action order based on the row order of the operation action name in the ancient building renovation operation sequence table, shift the disposal action order to the left of the construction node address width and then connect it to the construction node address to form the disposal action bit, and connect the disposal action bit to the end of the construction node address, and output the hidden part twin object string.

3. The method for dynamic management of the entire life cycle of ancient buildings based on digital and intelligent means according to claim 2, characterized in that: S2 includes: S21. Extract the construction node address of the target hidden part from the twin object string of the hidden part. Use the construction node address as the current node address. Read the building construction edge in the digital twin base map of the ancient building. The starting address is consistent with the current node address and the direction point is to the observation end. Write the end address of the building construction edge into the extrapolation path position. If the end address does not have the field status position, replace the current node address with the end address and continue reading. If the end address has the field status position, stop reading and output the observation component node address. S22. Based on the node address of the observed component, read the on-site status bit, and in the repair time sequence table, read the work items that the construction node address hit in reverse order of the work time. Shift the work order of the hit work items to the left by the work type bit width and then connect them to the work type bit to generate the recent intervention bit and output the observation acceptance record.

4. The method for dynamic management of the entire life cycle of ancient buildings based on digital and intelligent means according to claim 3, characterized in that: S2 also includes: S23. Read the structural layer number of the target hidden part and the structural layer number of the observation component node. Subtract the structural layer number of the observation component node from the structural layer number of the target hidden part to generate a layer difference value. Connect the sign bit of the layer difference value to the left of the absolute value bit of the layer difference value to generate a structural layer transfer bit. Write the field status bit, the near intervention bit, and the structural layer transfer bit into the status base in bit width order. S24. Count the number of unobserved state bits with empty values ​​in the state base and generate an unobserved count. Generate a binary expansion number starting from zero and incrementing by one until the binary expansion number reaches two to the power of the unobserved count minus one. Fill each binary expansion number into the unobserved state bit and output the candidate hidden state string.

5. The method for dynamic management of the entire life cycle of ancient buildings based on digital and intelligent means according to claim 4, characterized in that: S3 includes: S31. Based on the candidate hidden state string, read the candidate hidden state according to the candidate sequence number, write the candidate sequence number to the front of the free integer position, write the candidate hidden state to the back of the free integer position, shift the action position obtained by converting the candidate hidden state to the left by the width of the building cell and then connect it to the action cell, shift it to the left by the width of the entry cell and then connect it to the entry cell, and output the existence quantifier position. S32. Using the address of the node constructed from the hidden part of the target as the starting address, read the edge number of the active cell along the construction edge of the ancient building digital twin base map, accumulate the edge number to form a connection item, subtract the active cell from the entering cell to generate an order item, divide the action bit by the width of the building cell to generate a remainder item, and write the connection item, order item and remainder item into the Pressburg constraint string.

6. The method for dynamic management of the entire life cycle of ancient buildings based on digital and intelligent means according to claim 5, characterized in that: S3 also includes: S33. When performing Cooper Presburg arithmetic quantifier elimination on the Presburg constraint string, first rewrite the linear terms containing existential quantifiers as a unified term with an existential quantifier coefficient of one, then expand the remainder terms into remainder branches, then generate the lower bound of existential quantifiers with the connecting terms, generate the upper bound of existential quantifiers with the order terms, and replace the existential quantifiers with the lower bound plus the remainder branches. S34. After each remainder branch completes the replacement, delete the existing quantifier position, count the number of remaining existing quantifier positions, output the string without quantifier constraint when the number of remaining existing quantifier positions is zero, and return to S33 to process the next existing quantifier position when the number of remaining existing quantifier positions is not zero. S35. Substitute the candidate hidden state string back into the unqualified constraint string to generate the back-substitute action bit. Perform a bitwise XOR operation between the back-substitute action bit and the disposal action bit. When all bits of the XOR result are zero, compress the unqualified constraint string into an unqualified action signature. When the XOR result contains non-zero bits, write the candidate hidden state into the conflict elimination record.

7. The method for dynamic management of the entire life cycle of ancient buildings based on digital and intelligent means according to claim 6, characterized in that: S4 includes: S41. Based on the action signature without quantifiers, read the action signature pair in order of candidate hidden state, perform bitwise XOR on the action signature pair to generate candidate difference words, and merge the candidate difference words into the action difference word by bitwise OR. When all bits of the action difference word are zero, output no evidence collection management record. S42. When the action difference word contains a non-zero position, read the entrance position leading to the hidden part of the target from the digital twin base map of the ancient building as the left vertex, take the non-zero position in the action difference word as the right vertex, and write the bipartite graph edge when the building construction path from the entrance position to the hidden part of the target covers the non-zero position associated action grid. S43. For the bipartite graph edge, read from the entrance position along the building construction path to the target hidden part, write the component count data into the front part of the entrance cost word, shift the contact surface count data to the left and connect it to the entrance cost word, shift the uncover count data to the left and connect it to the entrance cost word, shift the enclosed space count data to the left and connect it to the entrance cost word, and convert the entrance cost word into a matching value.

8. The method for dynamic management of the entire life cycle of ancient buildings based on digital and intelligent means according to claim 7, characterized in that: S4 also includes: S44. When performing label initialization through Hungarian bipartite graph matching, the left vertex label is generated by reading the previous position in ascending order of the left vertex association matching cost, the right vertex label is written to zero, and the relaxation word is generated by subtracting the left and right vertex labels from the matching cost. The bipartite graph edges with a relaxation word of zero are written to the zero relaxation edge set. S45. Using the unmatched left vertex as the root node of the alternating tree, read the right vertex along the zero-relaxed edge set. When an unmatched right vertex is read, perform a matching flip along the path of the alternating tree. When a matched right vertex is read, connect the left vertex connected to the matched right vertex to the alternating tree. S46. When no unmatched right vertex is read in the alternating tree, read the relaxation words from the left vertex inside the alternating tree to the right vertex outside the alternating tree. Read the previous relaxation words in ascending order as the label rewriting words. Add the label rewriting words to the left vertex label and subtract the label rewriting words from the right vertex label. Then return to the zero relaxation edge set generation process until the number of matched right vertices reaches the number of non-zero bits in the action difference words. Output the evidence path record.

9. The method for dynamic management of the entire life cycle of ancient buildings based on digital and intelligent means according to claim 8, characterized in that: S5 includes: S51. Read the target hidden part address and path sequence according to the evidence collection path record, retrieve the evidence collection operation backfill item according to the path sequence, perform right truncation or left zero padding on the original observation code in the evidence collection operation backfill item according to the candidate hidden state bit width, generate the evidence collection result status word, and append the evidence collection result status word to the end of the target hidden part address, and output the evidence collection result record. S52. For the evidence collection result status word in the evidence collection result record, read the candidate hidden state in the candidate hidden state string in incrementing the candidate sequence number, perform bitwise XOR between the evidence collection result status word and the candidate hidden state, and then perform full bit OR operation on the bitwise XOR result. When the full bit OR result is zero, write the hit candidate sequence number and stop reading. When the full bit OR result is not zero, continue reading the next candidate hidden state. When the candidate hidden state reading ends and no hit candidate sequence number is written, output the miss flag. S53. Based on the hit candidate sequence number, read the hidden status of the hit candidate, locate the twin status slot in the digital twin base map of the ancient building with the target hidden part address, rewrite the status bit to be verified in the twin status slot to the hidden status of the hit candidate, and connect the evidence collection path record, evidence collection result status word and hit candidate sequence number to the building component product maintenance or repair management record according to the path order; based on the missing mark, rewrite the status bit to be verified in the twin status slot to the verification twin status, and output the pre-repair management decision record.

10. A dynamic management system for the entire life cycle of ancient buildings based on digital and intelligent means, used to implement the dynamic management method for the entire life cycle of ancient buildings based on digital and intelligent means as described in any one of claims 1-9, the system comprising a twin positioning module, a state expansion module, a quantifier elimination module, a path matching module, and a decision writing-back module, characterized in that: The twin positioning module is used to obtain the digital twin base map of the ancient building and the on-site work record before the renovation. It converts the hidden part number into the construction node address by the building part code, and converts the disposal actions in the work task into disposal action bits, and outputs the hidden part twin object string. The state expansion module, based on the target hidden part in the hidden part twin object string, extrapolates from the construction node address along the building construction edge to the observation component node, writes the on-site state of the observation component node, the recent intervention bit and the construction layer transmission bit into the state base segment by segment, performs binary expansion on the unobserved state bits, and outputs the candidate hidden state string. The quantifier elimination module is used to write the candidate hidden state into the free integer position through the candidate hidden state string, and to connect the action position through the action cell and the entry cell to form the existence quantifier position. The Cooper Pressburg arithmetic quantifier elimination is performed by the connector, the order, and the remainder, and the output quantifier-free action signature is output. The path matching module is used to generate action difference words by bitwise XORing the action signature without quantifiers. When all bits of the action difference word are zero, no evidence collection management record is output. When the action difference word contains non-zero bits, an evidence collection bipartite graph is constructed with the entry bit and the non-zero bits, and Hungarian bipartite graph matching is performed with the entry cost word to output the evidence collection path record. The decision write-back module is used to generate evidence collection result records based on the evidence collection path records, perform position matching between the evidence collection result records and candidate hidden state strings, and write back the twin state of the target hidden part and the maintenance or repair management record of the building component product when a match is found, and write back the review twin state when a match is not found, and output the management decision record before repair.

Images