Engineering cost data security management method based on block chain evidence storage

Abstract

The invention discloses an engineering cost data security management method based on block chain evidence storage. The method comprises the following steps: obtaining an engineering cost file and extracting meta-information for encrypted storage; unifying and normalizing the lines of the list and constructing a line-level Merkle tree; constructing a grouping Merkle tree and a document level Merkle tree for the row level Hash grouping; generating a differential event based on the new and old version list rows and constructing a cost relationship knowledge graph; legal differences are screened to construct a difference Merkle tree, and a double-layer evidence storage structure is formed; and summarizing constraint and signature information to generate chained anchoring and chaining, thereby realizing safety verification and traceable management of the project cost data. According to the method, block chain evidence storage and constraint satisfaction type knowledge graph are fused, credible evidence storage and traceable verification of multi-version difference of project cost data are achieved, and the method has the advantages of being high in consistency, fine in verification and high in management reliability.

Description

Technical Field

[0001] This invention relates to the field of information technology in engineering management, and in particular to a method for secure management of engineering cost data based on blockchain-based evidence storage. Background Technology

[0002] Existing engineering cost data management typically relies on engineering cost documents and their attachments. Archiving and version iterations are conducted around the entry, modification, approval, measurement, and settlement of bill of quantities records, and tamper-proofing is enhanced through database auditing, electronic signatures, or on-chain notarization. Some solutions generate summaries for documents or fields and upload them to the blockchain, or construct Merkle trees using hash aggregation to support integrity verification and traceability. Other solutions combine off-chain encrypted storage with index mapping to achieve searchable and verifiable evidence, meeting the needs for data traceability and accountability in cross-entity collaboration.

[0003] However, in engineering cost scenarios involving multiple version evolutions and multi-entity confirmations, existing technologies often struggle to simultaneously achieve a closed-loop integration of row-level location verification, group verification based on project ownership information / contract segment ownership information / cost stage information, and differential verification. This can easily lead to issues such as being able to verify only the entire document or a coarse-grained summary, and being unable to generate row-level and group-level proofs. More importantly, differential comparison results typically lack strong binding with structured consistency rules, failing to introduce constraint-satisfying knowledge graph consistency judgment and tree entry gating before differential events are added to the chain. This results in differential events that do not meet hard constraints still being stored as evidence, thereby weakening the support of differential proofs and constraint consistency proofs for traceability management.

[0004] Therefore, how to provide a secure management method for engineering cost data based on blockchain-based evidence storage is a problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0005] One objective of this invention is to propose a secure management method for engineering cost data based on blockchain notarization. This invention integrates blockchain notarization with constraint-satisfying knowledge graph methods to achieve reliable notarization and traceable verification of multi-version differential engineering cost data, which has the advantages of strong consistency, fine verification, and high management reliability.

[0006] A method for secure management of engineering cost data based on blockchain-based notarization according to an embodiment of the present invention includes the following steps:

[0007] Obtain the engineering cost documents and related attachments, extract the document metadata summary and generate a set of engineering cost list row records, perform encrypted storage on the engineering cost documents and related attachments, and form an off-chain secure library and index mapping set;

[0008] Based on the rule version number and view label, the set of row records in the project cost list is uniformly normalized, and a row-level hash set is generated as the leaf node to construct a row-level Merkle tree, thus obtaining the row-level root;

[0009] The row-level hash set is grouped according to the preset cost grouping key, and a grouped Merkle tree is constructed to obtain the group root set. The document-level Merkle tree is then constructed using the leaf nodes to obtain the grouped document root.

[0010] Perform differential comparison on the current version and the parent version of the project cost list row record set to generate a differential event list. Combine the off-chain security library, index mapping set and document meta information digest to construct a project cost relationship knowledge graph.

[0011] Constraint-satisfaction knowledge graph consistency determination is performed. Based on the constraint satisfaction results, tree gating is performed on the differential event list to generate a tree-enterable differential event list. A differential Merkle tree is constructed to obtain the differential root, and a cost group differential two-layer Merkle evidence storage structure is established.

[0012] Generate constraint satisfaction digests and signature set digests, combine cost grouping differential two-layer Merkle notarization structure and document metadata digest to generate anchor hashes to obtain on-chain anchors, write them to the consortium blockchain to form notarization records, and output row proofs, group proofs, differential proofs and constraint consistency proofs to complete the security verification and traceability management of engineering cost data.

[0013] Optionally, the generation of the project cost list row record set and the off-chain security library specifically includes:

[0014] Obtain the project cost documents and related attachments, and categorize them according to document source, business stage, and version to form a set of documents to be processed;

[0015] The set of files to be processed is subjected to format unification and structured parsing to obtain a set of normalized fields;

[0016] The document metadata item set is extracted based on the normalized field set, normalized encoding and sequence solidification are performed to generate the document metadata sequence, and the summary generation operator is used to generate the summary to obtain the document metadata summary;

[0017] Each list row record is solidified into a row-level structured entry based on the standardized field set, and then aggregated to generate a list of project cost list row records;

[0018] The project cost documents and related attachments are encrypted and stored, generating a set of ciphertext files and a set of key encapsulations, which are then written to an off-chain security library.

[0019] A set of off-chain security repository location information is generated based on the set of encrypted files. An association mapping is established between the set of off-chain security repository location information and the set of engineering cost list line records to generate an index mapping set.

[0020] Optionally, the generation of the row-level root specifically includes:

[0021] Based on the rule version number and view label, the engineering cost list row record set is processed by field selection, field order fixing and field value normalization to obtain a row-level normalized field set;

[0022] For each row-level normalized field entry in the row-level normalized field set, write the rule version number and view label in the preset position to generate a row-level normalized sequence. Then, use the digest generation operator to perform row-level hash generation to obtain row-level hash values ​​and aggregate them to generate a row-level hash set.

[0023] The row-level hash set is sorted and solidified based on the unique locator key of the row-level structured entry, and then written into the leaf node sequence to generate the leaf node set;

[0024] A row-level Merkle tree is constructed using the set of leaf nodes as leaf nodes. The parent node hash is generated by sequentially performing binary combinations on adjacent leaf nodes and using a digest generation operator. The nodes of each level are generated iteratively. When the number of nodes in a certain level is odd, the last node of that level is copied sequentially to make up the gap before performing binary combinations again, until a unique row-level root is generated.

[0025] Optionally, the generation of the grouped document root specifically includes:

[0026] Obtain the row-level hash set and row-level root, and read the project ownership information, contract section ownership information and cost stage information corresponding to each row-level hash value from the project cost list row record set to form a row-level hash-cost grouping key association set;

[0027] Based on the preset cost grouping key, the row-level hash-cost grouping key association set is grouped and merged to obtain the grouped row-level hash subset;

[0028] For each group's row-level hash subset, perform intra-group sorting and solidification to generate a group leaf node sequence, and use these leaf nodes to construct a group Merkle tree to obtain the group root set;

[0029] Establish a group root-cost group key binding relationship between the group root set and the corresponding cost group key value, and generate a group root registration table;

[0030] Based on the preset cost grouping key order rules, the leaf node order of the grouping root registration table is fixed to generate the grouping root leaf node sequence, and the document-level Merkle tree is constructed as the leaf nodes to obtain the grouping document root.

[0031] Optionally, the generation of the engineering cost relationship knowledge graph specifically includes:

[0032] Get the set of project cost list row records for the current version and the parent version, perform row-level alignment, and generate a version alignment mapping set;

[0033] Based on the version alignment mapping set, a differential comparison is performed on the current version and the parent version of the project cost list row record set to generate a differential event list;

[0034] Write each differential event in the differential event list into its corresponding unique location key, parent version attribution information, current version attribution information, and differential field set to form a structured entry set of differential events;

[0035] Perform event normalization processing on the structured set of differential events to generate a normalized list of differential events;

[0036] A knowledge graph of engineering cost relationships is constructed based on a normalized differential event list, an off-chain security library, an index mapping set, and a document meta-information summary, generating a knowledge graph entity set and a knowledge graph relationship set.

[0037] The knowledge graph entity set and knowledge graph relation set are subjected to consistent encoding and sequential solidification to generate a knowledge graph structure sequence. The off-chain security library location information corresponding to the knowledge graph entity set in the document meta-information digest and index mapping set is written into the knowledge graph structure sequence, and the engineering cost relation knowledge graph is output.

[0038] Optionally, the generation of the cost grouping differential two-layer Merkle evidence storage structure specifically includes:

[0039] Obtain the knowledge graph of engineering cost relationships, extract the knowledge graph entity set and knowledge graph relationship set corresponding to the differential event list, and form the differential consistency judgment input graph.

[0040] Based on a preset set of hard constraints, the differential consistency judgment input graph is subjected to a constraint-satisfaction knowledge graph consistency judgment, generating a set of constraint-satisfaction results.

[0041] Based on the constraint satisfaction result set, perform tree-entry gating processing on the difference event list to generate a tree-entry difference event list;

[0042] Perform event-level normalization and sequence solidification on the list of possible differential events to generate a normalized differential event sequence;

[0043] Construct a difference Merkle tree using the normalized difference event sequence as leaf nodes to obtain the difference root;

[0044] A cost grouping differential two-layer Merkle evidence storage structure is established based on the grouped root set, grouped document root, and differential root.

[0045] Optionally, the generation of the engineering cost data security verification and traceability management specifically includes:

[0046] Obtain the cost grouping differential two-layer Merkle evidence storage structure, document metadata summary, constraint satisfaction result set, and multi-subject signature set involved in engineering cost management; generate constraint satisfaction summary based on constraint satisfaction result set, and generate signature set summary based on multi-subject signature set;

[0047] The constraint satisfaction summary, signature set summary, document metadata summary, and cost grouping differential two-layer Merkle evidence storage structure are solidified to generate an anchored input sequence.

[0048] An anchor hash is generated by using a digest generation operator on the anchor input sequence to obtain the on-chain anchor. An on-chain anchor-index mapping and binding relationship is established between the on-chain anchor and the off-chain security library location information set to generate an on-chain anchor registration table.

[0049] The on-chain anchor, cost group differential two-layer Merkle notarization structure, document meta-information digest, constraint satisfaction digest, signature set digest, and on-chain anchor registration table are written into the consortium blockchain to form notarization records, and a notarization record index information set is generated.

[0050] Row proofs are generated based on row-level Merkle trees, group proofs are generated based on grouped Merkle trees, difference proofs are generated based on difference Merkle trees, and constraint consistency proofs are generated based on the constraint satisfaction result set and the difference consistency determination input graph.

[0051] During the verification and traceability phase, the system receives the line records or differential events of the engineering cost list to be verified, locates the evidence record index information set based on the on-chain anchor registration table, and reads the cost group differential two-layer Merkle evidence storage structure, document meta information digest, constraint satisfaction digest, and signature set digest corresponding to the on-chain anchor. The system then performs path recalculation and consistency verification using line proof, group proof, differential proof, and constraint consistency proof, outputs the verification conclusion, and completes the traceability management of engineering cost data when the consistency verification passes.

[0052] The beneficial effects of this invention are:

[0053] Through the above technical solutions, this invention achieves end-to-end secure evidence storage and verifiable closed-loop management in scenarios involving multiple versions of engineering cost data evolution and multi-entity collaborative management, from the list row level, cost group level to the differential event level. Based on the unified and standardized processing of rule version numbers and view labels, engineering cost list rows possess a consistent row-level hash expression across different business stages and version environments. Combined with the hierarchical structure of row-level Merkle trees and group Merkle trees, it enables precise location verification of individual list rows, specific cost groups, and complete documents. This avoids the problem of existing technologies only being able to perform file-level or coarse-grained integrity checks, thereby significantly improving the verifiability and operational reliability of engineering cost data during auditing, review, and accountability processes.

[0054] Furthermore, by introducing a constraint-satisfaction-based knowledge graph consistency determination before differential events are added to the chain, and using it as the sole legitimate entry point for differential events into the differential Merkle tree, this invention structurally binds engineering cost business rules, version inheritance relationships, and approval closed-loop requirements to the blockchain evidence storage structure. This ensures that only differential events that satisfy all hard constraints can form on-chain anchors. The on-chain anchors generated by combining multi-principal signature set digests, constraint satisfaction digests, and the cost group differential two-layer Merkle evidence storage structure not only enhance the credibility of differential proofs and constraint consistency proofs but also support rapid verification and traceability across versions and groups, thereby improving the overall security, compliance, and traceability management capabilities of engineering cost data management. Attached Figure Description

[0055] The accompanying drawings are provided to further illustrate the invention and form part of the specification. They are used in conjunction with embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings:

[0056] Figure 1 The flowchart shows a method for secure management of engineering cost data based on blockchain evidence storage proposed in this invention.

[0057] Figure 2 This is a schematic diagram of the row-level Merkle tree and grouped document Merkle tree structure of the engineering cost data security management method based on blockchain notarization proposed in this invention.

[0058] Figure 3 This is a schematic diagram of the cost grouping differential two-layer Merkle notarization structure of the engineering cost data security management method based on blockchain notarization proposed in this invention. Detailed Implementation

[0059] The present invention will now be described in further detail with reference to the accompanying drawings. These drawings are simplified schematic diagrams, illustrating only the basic structure of the invention, and therefore only show the components relevant to the invention.

[0060] refer to Figures 1-3 A method for secure management of engineering cost data based on blockchain-based notarization includes the following steps:

[0061] Obtain the engineering cost documents and related attachments, extract the document metadata summary and generate a set of engineering cost list row records, perform encrypted storage on the engineering cost documents and related attachments, and form an off-chain secure library and index mapping set;

[0062] Based on the rule version number and view label, the set of row records in the project cost list is uniformly normalized, and a row-level hash set is generated as the leaf node to construct a row-level Merkle tree, thus obtaining the row-level root;

[0063] The row-level hash set is grouped according to the preset cost grouping key, and a grouped Merkle tree is constructed to obtain the group root set. The document-level Merkle tree is then constructed using the leaf nodes to obtain the grouped document root.

[0064] Perform differential comparison on the current version and the parent version of the project cost list row record set to generate a differential event list. Combine the off-chain security library, index mapping set and document meta information digest to construct a project cost relationship knowledge graph.

[0065] Constraint-satisfaction knowledge graph consistency determination is performed. Based on the constraint satisfaction results, tree gating is performed on the differential event list to generate a tree-enterable differential event list. A differential Merkle tree is constructed to obtain the differential root, and a cost group differential two-layer Merkle evidence storage structure is established.

[0066] Generate constraint satisfaction digests and signature set digests, combine cost grouping differential two-layer Merkle notarization structure and document metadata digest to generate anchor hashes to obtain on-chain anchors, write them to the consortium blockchain to form notarization records, and output row proofs, group proofs, differential proofs and constraint consistency proofs to complete the security verification and traceability management of engineering cost data.

[0067] In this embodiment, the generation of the engineering cost list row record set and the off-chain security library specifically includes:

[0068] Obtain the project cost documents and related attachments, and categorize them according to document source, business stage, and version to form a set of documents to be processed;

[0069] The set of files to be processed is subjected to format unification and structured parsing to obtain a set of normalized fields;

[0070] The structured parsing includes extracting structured fields from the list page, change page, visa page, measurement page, and settlement page, and performing missing item removal and conflict item merging;

[0071] The document metadata item set is extracted based on the normalized field set, normalized encoding and sequence solidification are performed to generate the document metadata sequence, and the summary generation operator is used to generate the summary to obtain the document metadata summary;

[0072] The document metadata sequence is obtained by splicing together document identity information, project ownership information, contract section ownership information, cost stage information, version ownership information and associated attachment list information in a preset order;

[0073] The summary generation process involves converting a document metadata sequence into a continuous byte stream under a fixed character set, dividing the byte stream into blocks according to a preset block length, performing a compression mapping operation on each block to generate an intermediate summary value, concatenating and combining the intermediate summary values ​​according to the block order, and repeatedly performing the compression mapping operation until a final summary result with a fixed length is obtained. The compression mapping operation employs irreversible bit mixing, shifting, and nonlinear substitution rules.

[0074] Each list row record is solidified into a row-level structured entry based on the standardized field set, and then aggregated to generate a list of project cost list row records;

[0075] The row-level structured entries are organized by page type, table area and row number, the boundaries of the list rows to which the fields belong are identified and the list rows across pages and table areas are aligned and merged to form a list row candidate set. A row-level structured entry template is built for each list row candidate, the corresponding set of field key-value pairs is written into the row-level structured entry template according to the preset field mapping rules, and missing items are removed and conflict items are merged to obtain row-level structured entries with consistent fields. A unique location key is generated based on the project ownership information, contract section ownership information, list row location information and the core fields of the list row, and written into the row-level structured entry.

[0076] The project cost documents and related attachments are encrypted and stored, generating a set of ciphertext files and a set of key encapsulations, which are then written to an off-chain security library.

[0077] The encrypted file set is generated by using an encryption storage operator and a data encryption key to encrypt the plaintext content of the engineering cost documents and related attachments; the key encapsulation set is generated by using a key encapsulation operator and an access control key to encapsulate the data encryption key.

[0078] A set of off-chain security repository location information is generated based on the set of encrypted files. An association mapping is established between the set of off-chain security repository location information and the set of engineering cost list line records to generate an index mapping set.

[0079] In this embodiment, the generation of the row-level root specifically includes:

[0080] Based on the rule version number and view label, the engineering cost list row record set is processed by field selection, field order fixing and field value normalization to obtain a row-level normalized field set;

[0081] The field selection process predefines a unique field closure configuration table for each rule version number, limiting the set of field names participating in row-level hash generation. Based on the current rule version number, the fields limited by the field closure configuration table are selected from the set of engineering cost list row records to form a set of fields participating in hashing. The field order fixing is defined by predefining a field order description table for each view label, limiting the concatenation order of the fields participating in hashing in the row-level normalized sequence. Based on the current view label, the field set is concatenated according to the order specified in the field order description table to generate a unique row-level normalized sequence. The field value normalization processing includes performing unit consistency and decimal place fixing for numeric fields, character set unification and whitespace normalization for text fields, and format fixing and time zone normalization for date fields.

[0082] For each row-level normalized field entry in the row-level normalized field set, write the rule version number and view label in the preset position to generate a row-level normalized sequence. Then, use the digest generation operator to perform row-level hash generation to obtain row-level hash values ​​and aggregate them to generate a row-level hash set.

[0083] The row-level hash set is sorted and solidified based on the unique locator key of the row-level structured entry, and then written into the leaf node sequence to generate the leaf node set;

[0084] Each leaf node in the leaf node set establishes a location binding relationship with the unique locator key of its corresponding list row entry, which is used to locate the path position of the leaf node in the row-level Merkle tree.

[0085] A row-level Merkle tree is constructed using the set of leaf nodes as leaf nodes. The parent node hash is generated by sequentially performing binary combinations on adjacent leaf nodes and using a digest generation operator. The nodes of each level are generated iteratively. When the number of nodes in a certain level is odd, the last node of that level is copied sequentially to complete the binary combination, until a unique row-level root is generated.

[0086] The parent node hash generation is achieved by converting the hashes of the left and right child nodes into a continuous byte stream in a preset order and concatenating them to form a combined input. The digest generation operator is then applied to the combined input to obtain the parent node hash.

[0087] In this embodiment, the generation of the grouped document root specifically includes:

[0088] Obtain the row-level hash set and row-level root, and read the project ownership information, contract section ownership information and cost stage information corresponding to each row-level hash value from the project cost list row record set to form a row-level hash-cost grouping key association set;

[0089] Based on the preset cost grouping key, the row-level hash-cost grouping key association set is grouped and merged to obtain the grouped row-level hash subset;

[0090] The preset cost grouping key is generated by combining the existing project ownership information, contract segment ownership information, and cost stage information in the project cost list row record set according to preset splicing rules. The project ownership information is used to represent the project object to which the list row belongs, the contract segment ownership information is used to represent the contract execution unit to which the list row belongs, and the cost stage information is used to represent the cost business stage to which the list row belongs. The cost grouping key is used as the sole criterion for determining row-level hash grouping, limiting the row-level hash to enter the corresponding group Merkle tree. The grouping is performed by extracting the corresponding cost grouping key value for each row-level hash value, and writing the row-level hash values ​​with the same cost grouping key value into the same group row-level hash subset, and outputting a list of group row-level hash subsets.

[0091] For each group's row-level hash subset, perform intra-group sorting and solidification to generate a group leaf node sequence, and use these leaf nodes to construct a group Merkle tree to obtain the group root set;

[0092] The sorting and solidification within the group is achieved by sorting and solidifying the row-level hash subset of the group according to the unique location key of the row-level structured entry and writing it sequentially into the group leaf node sequence. The group Merkle tree is generated by sequentially combining adjacent group leaf nodes into binary combinations and using a digest generation operator to generate the parent node hash, iterating to generate group nodes at each level until the group root of the group Merkle tree is generated.

[0093] Establish a group root-cost group key binding relationship between the group root set and the corresponding cost group key value, and generate a group root registration table;

[0094] Based on the preset cost grouping key order rules, the leaf node order of the grouping root registration table is fixed to generate the grouping root leaf node sequence, and the document-level Merkle tree is constructed as the leaf nodes to obtain the grouping document root;

[0095] The leaf node order solidification is achieved by sorting and solidifying the cost grouping key values ​​in the grouping root registration table, and writing them sequentially into the corresponding grouping roots according to the sorted cost grouping key values. The document-level Merkle tree is generated by sequentially combining adjacent leaf nodes into binary combinations and using a digest generation operator to generate a parent node hash, iterating through each level of nodes until a unique grouped document root is generated.

[0096] In this embodiment, the generation of the engineering cost relationship knowledge graph specifically includes:

[0097] Get the set of project cost list row records for the current version and the parent version, perform row-level alignment, and generate a version alignment mapping set;

[0098] The version alignment mapping set includes the correspondence between parent version list row records and current version list row records, missing correspondence, and new correspondence;

[0099] Based on the version alignment mapping set, a differential comparison is performed on the current version and the parent version of the project cost list row record set to generate a differential event list;

[0100] The differential comparison includes performing field-level comparison on the list row records of the corresponding relationship and generating change events, generating deletion events for missing corresponding relationships, and generating addition events for newly added corresponding relationships;

[0101] Write each differential event in the differential event list into its corresponding unique location key, parent version attribution information, current version attribution information, and differential field set to form a structured entry set of differential events;

[0102] Perform event normalization processing on the structured set of differential events to generate a normalized list of differential events;

[0103] The event normalization process includes performing field selection, field order fixing, and field value normalization on the differential field set, and writing the rule version number and view label into each differential event structured entry;

[0104] A knowledge graph of engineering cost relationships is constructed based on a normalized differential event list, an off-chain security library, an index mapping set, and a document meta-information summary, generating a knowledge graph entity set and a knowledge graph relationship set.

[0105] The knowledge graph entity set includes document entities, version entities, list line entities, differential event entities, and off-chain security library object entities; the knowledge graph relationship set includes version inheritance relationships, list line attribution relationships, differential association relationships, document-attachment reference relationships, and entity-off-chain security library location relationships.

[0106] The knowledge graph entity set and knowledge graph relation set are subjected to consistent encoding and sequential solidification to generate a knowledge graph structure sequence. The off-chain security library location information corresponding to the knowledge graph entity set in the document meta-information digest and index mapping set is written into the knowledge graph structure sequence, and the engineering cost relation knowledge graph is output.

[0107] In this embodiment, the generation of the cost grouping differential two-layer Merkle evidence storage structure specifically includes:

[0108] Obtain the knowledge graph of engineering cost relationships, extract the knowledge graph entity set and knowledge graph relationship set corresponding to the differential event list, and form the differential consistency judgment input graph.

[0109] The differential consistency determination input graph includes differential event entities, list row entities, version entities, and differential association relationships, version inheritance relationships, and list row attribution relationships between entities;

[0110] Based on a preset set of hard constraints, the differential consistency judgment input graph is subjected to a constraint-satisfaction knowledge graph consistency judgment, generating a set of constraint-satisfaction results.

[0111] The set of hard constraints is a closed list, including that each list row entity is only allowed to belong to a unique contract segment attribution information, the parent version attribution information of each differential event entity must point to an existing and unterminated version entity in the engineering cost relationship knowledge graph, the deletion event must have a corresponding replacement relationship or invalidation relationship in the engineering cost relationship knowledge graph, and each differential event entity must have a complete approval relationship closed loop in the engineering cost relationship knowledge graph.

[0112] The constraint-satisfaction type knowledge graph consistency determination examines the satisfaction of the hard constraint set one by one in the differential consistency determination input graph and outputs the determination result.

[0113] Based on the constraint satisfaction result set, perform tree-entry gating processing on the difference event list to generate a tree-entry difference event list;

[0114] The tree entry gate is the only legitimate entry point for a differential event to enter the differential Merkle tree. A differential event is allowed to be written into the tree entry differential event list only when it is marked as satisfying all hard constraints in the constraint satisfaction result set. Differential events that do not satisfy any hard constraints are excluded from the tree entry differential event list.

[0115] Perform event-level normalization and sequence solidification on the list of possible differential events to generate a normalized differential event sequence;

[0116] The event-level normalization includes performing field selection, field order fixing, and field value normalization processing on the event type, associated list row location information, parent version attribution information, and current version attribution information of differential events;

[0117] Construct a difference Merkle tree using the normalized difference event sequence as leaf nodes to obtain the difference root;

[0118] The differential Merkle tree generates differential nodes at each level by sequentially combining adjacent normalized differential event sequences and generating parent node hashes using a digest generation operator. When the number of nodes at a certain level is odd, the last node of that level is copied sequentially to complete the binary combination, until a unique differential root is generated.

[0119] A cost grouping differential two-layer Merkle evidence storage structure is established based on the grouped root set, grouped document root, and differential root;

[0120] The cost grouping differential two-layer Merkle evidence storage structure includes a two-layer Merkle structure with the group document root as the upper-layer structure anchor point and the differential root as the lower-layer differential anchor point, which is used to structurally bind the group-level integrity proof and the differential-level legality proof.

[0121] In this embodiment, the generation of the engineering cost data security verification and traceability management specifically includes:

[0122] Obtain the cost grouping differential two-layer Merkle evidence storage structure, document metadata summary, constraint satisfaction result set, and multi-subject signature set involved in engineering cost management; generate constraint satisfaction summary based on constraint satisfaction result set, and generate signature set summary based on multi-subject signature set;

[0123] The multi-entity signature set for participating in project cost management is a set of signature results formed by multiple business entities participating in the project cost management signing and confirming the current version of the project cost list row record set, the differential event list, or their summary results during the generation of the current version of the project cost document and the confirmation of differential events.

[0124] The constraint satisfaction summary includes hard constraint set identifiers, differential event list identifiers, constraint satisfaction result set identifiers, and non-satisfaction item location information set; the signature set summary includes signature subject identifier set, signature order solidification result, and multi-subject signature set.

[0125] The constraint satisfaction summary, signature set summary, document metadata summary, and cost grouping differential two-layer Merkle evidence storage structure are solidified to generate an anchored input sequence.

[0126] The abstract input order solidification is achieved by concatenating the constraint satisfaction abstract, signature set abstract, and document meta information abstract according to a preset abstract order rule, and writing the grouped document roots and difference roots in the cost group differential double-layer Merkle evidence storage structure into the anchored input sequence according to a preset structure order rule, thus obtaining an anchored input sequence with a unique order.

[0127] An anchor hash is generated by using a digest generation operator on the anchor input sequence to obtain the on-chain anchor. An on-chain anchor-index mapping and binding relationship is established between the on-chain anchor and the off-chain security library location information set to generate an on-chain anchor registration table.

[0128] The on-chain anchor registration table is used to locate the corresponding cost group differential two-layer Merkle evidence storage structure, document meta information digest, constraint satisfaction digest and signature set digest in the subsequent verification stage, using the on-chain anchor as the unique index.

[0129] The on-chain anchor, cost group differential two-layer Merkle notarization structure, document meta-information digest, constraint satisfaction digest, signature set digest, and on-chain anchor registration table are written into the consortium blockchain to form notarization records, and a notarization record index information set is generated.

[0130] The evidence storage record index information set includes on-chain anchors, rule version numbers, view tags, cost grouping key value sets, and parent version attribution information, which are used to support subsequent cross-version tracing and differential verification;

[0131] Row proofs are generated based on row-level Merkle trees, group proofs are generated based on grouped Merkle trees, difference proofs are generated based on difference Merkle trees, and constraint consistency proofs are generated based on the constraint satisfaction result set and the difference consistency determination input graph.

[0132] The row proof includes a unique location key, the corresponding row-level hash value, and its path information to the row-level root; the group proof includes the cost group key value, the corresponding group root, and its path information to the group document root; the difference proof includes the differential event structured entry and its path information to the difference root; and the constraint consistency proof includes a hard constraint set identifier, a differential event structured entry identifier, a constraint satisfaction result set identifier, and a non-satisfied item location information set.

[0133] During the verification and traceability phase, the system receives the line records or differential events of the engineering cost list to be verified, locates the evidence record index information set based on the on-chain anchor registration table, and reads the cost group differential two-layer Merkle evidence storage structure, document meta information digest, constraint satisfaction digest, and signature set digest corresponding to the on-chain anchor. The system then performs path recalculation and consistency verification using line proof, group proof, differential proof, and constraint consistency proof, outputs the verification conclusion, and completes the traceability management of engineering cost data when the consistency verification passes.

[0134] Example 1:

[0135] To verify the feasibility of this invention in practice, it was applied to the whole-process cost management scenario of a large-scale infrastructure project. This project spans multiple construction phases and involves various cost documents, including design, construction, changes, and settlement documents. These documents are scattered, frequently updated, and involve complex adjustments, replacements, and cancellations at the list line level. This has long resulted in difficulties in verifying cost data consistency, unclear responsibility boundaries, and high costs for post-project traceability. Traditional methods, relying mainly on manual comparison and centralized archiving, are insufficient to meet the requirements of refined supervision and cross-entity collaborative management.

[0136] In this scenario, the engineering cost documents and their associated attachments are uniformly integrated into the method flow of this invention during the formation stage. The document content is structurally parsed to generate a set of engineering cost list line records, and the document metadata digest is extracted simultaneously. All original files are encrypted and stored in an off-chain security repository, and an index mapping relationship is established between the list line records and the storage location of the security repository. Subsequently, the list line records are uniformly normalized according to the rule version number and view label to ensure a stable and consistent row-level hash result under different business perspectives and rule versions. Furthermore, a Merkle structure at the row and group levels is constructed to lay the foundation for subsequent integrity verification.

[0137] When a project enters the change or settlement phase, the current version of the inventory list records is automatically aligned and differentially compared with the previous version, generating a differential event list containing additions, deletions, and changes. These differential events are not directly used for evidence preservation; instead, they are used in conjunction with document metadata summaries and off-chain security library index information to construct a knowledge graph of project cost relationships, clearly defining inventory list lines, versions, differential events, and their relationships. Based on this, a constraint-satisfaction-based consistency determination mechanism is introduced, verifying each rule regarding the unique attribution of inventory list lines, the validity of version inheritance, deletion substitution relationships, and the integrity of the approval chain. Only when a differential event satisfies all hard constraints is it allowed to proceed to the subsequent differential evidence preservation process.

[0138] This constraint gating mechanism effectively prevents issues such as unapproved changes, incorrect version pointers, or duplicate list row attributions from being written into the evidence storage structure in practical applications. Qualified differential events are normalized and a differential Merkle structure is constructed, forming a two-layer evidence storage relationship with the Merkle results at the grouped document level. Subsequently, the constraint satisfaction digest, the signature set digest formed by multi-party confirmation, and the document metadata digest are combined to generate an anchor hash, which is then written into the consortium blockchain to complete the evidence storage.

[0139] During actual operation, when project cost management personnel verify historical bill of quantities adjustments at a specific construction node, they only need to provide the corresponding bill of quantities row or differential event. Based on the on-chain anchor, the relevant evidence can be quickly located, and verification can be completed through row proof, group proof, differential proof, and constraint consistency proof. Application results show that this method significantly reduces the manual burden of cross-version verification, enabling clear and verifiable traceability of the cost data's formation process, evolution path, and legality status. This provides a stable and reliable means of data security and accountability confirmation for project cost management.

[0140] Table 1. Comparison of Overall Effectiveness of Engineering Cost Data Security Management Methods

[0141] Comparison indicators Traditional cost management methods Blockchain-based row-level evidence storage method Method of the present invention List row-level differential recognition accuracy (%) 90.4 94.1 98.6 Success rate of cross-version cost tracing (%) 88.7 93.5 99.2 Percentage of illegal or incomplete changes to be archived (%) 7.8 3.2 0.9 Completeness of traceability of multi-entity liability (%) 82.3 91.6 96.8 Cost consistency dispute rate (times / 100 versions) 6.4 2.7 0.8 Average processing time (seconds) for a single cross-version verification 12.6 6.8 3.9

[0142] The accuracy of differential identification at the row level in the bill of quantities shows that traditional cost management methods mainly rely on manual comparison or coarse-grained field verification. Under complex version evolution conditions, these methods are easily affected by missing fields and structural changes, resulting in a low accuracy rate. While blockchain-based row-level evidence storage methods improve differential identification capabilities to some extent by introducing a row-level hashing mechanism, they still lack structural constraints on the legality of differentials. This invention, through row-level normalization driven by rule version numbers and view tags, combined with the structured expression of differential events, ensures high consistency in differential comparisons at the field level, thereby significantly improving the stability of differential identification.

[0143] Regarding the success rate of cross-version cost tracing, traditional methods are prone to path breaks or missing connections after multiple changes, resulting in some historical versions not being fully traceable. While blockchain notarization enhances version immutability, the lack of a unified standard for determining the legitimacy of differential events still leads to interruptions in the tracing process. This invention constructs a two-layer Merkle notarization structure for cost grouping and differential notarization, structurally binding group-level integrity with differential-level legitimacy. This ensures that the cross-version tracing process always has a clear path location basis, significantly improving the tracing success rate.

[0144] Regarding the proportion of illegal or incomplete changes entering the archive, traditional cost management models mainly rely on manual review or post-event verification, which is insufficient to promptly prevent changes that do not comply with business rules. While row-level evidence storage methods can record change results, they lack pre-emptive constraints on the rationality of changes. This invention introduces a constraint-satisfaction-based knowledge graph consistency determination, using whether a differential event satisfies hard constraints as the sole legitimate entry point into the differential Merkle tree. This systematically excludes changes that do not meet constraints at the technical level, thereby effectively reducing the proportion of illegal or incomplete changes entering the archive.

[0145] The comparison results of the traceability completeness of multi-entity responsibility show that, in multi-party collaborative management scenarios, traditional methods often result in responsibility chains being scattered across different documents and systems, making it difficult to form a unified and continuous responsibility mapping relationship. This invention, when generating on-chain anchors, uniformly anchors the multi-entity signature set digest, constraint satisfaction digest, and cost grouping differential two-layer Merkle evidence storage structure, ensuring that each version confirmation and differential confirmation has a verifiable source of responsibility, thereby significantly enhancing the completeness of multi-entity responsibility traceability.

[0146] Furthermore, in terms of the incidence of cost consistency disputes and verification processing efficiency, this invention uses a combination of row proofs, group proofs, differential proofs and constraint consistency proofs to enable the verification process to complete path recalculation and consistency verification without loading all data. This not only reduces the probability of disputes but also effectively shortens the processing time for cross-version verification, further improving the overall operational efficiency of engineering cost data security management.

[0147] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A method for secure management of engineering cost data based on blockchain-based notarization, characterized in that, Includes the following steps: Obtain the engineering cost documents and related attachments, extract the document metadata summary and generate a set of engineering cost list row records, perform encrypted storage on the engineering cost documents and related attachments, and form an off-chain secure library and index mapping set; Based on the rule version number and view label, the set of row records in the project cost list is uniformly normalized, and a row-level hash set is generated as the leaf node to construct a row-level Merkle tree, thus obtaining the row-level root; The row-level hash set is grouped according to the preset cost grouping key, and a grouped Merkle tree is constructed to obtain the group root set. The document-level Merkle tree is then constructed using the leaf nodes to obtain the grouped document root. Perform differential comparison on the current version and the parent version of the project cost list row record set to generate a differential event list. Combine the off-chain security library, index mapping set and document meta information digest to construct a project cost relationship knowledge graph. Constraint-satisfaction knowledge graph consistency determination is performed. Based on the constraint satisfaction results, tree gating is performed on the differential event list to generate a tree-enterable differential event list. A differential Merkle tree is constructed to obtain the differential root, and a cost group differential two-layer Merkle evidence storage structure is established. Generate constraint satisfaction digests and signature set digests, combine cost grouping differential two-layer Merkle notarization structure and document metadata digest to generate anchor hashes to obtain on-chain anchors, write them to the consortium blockchain to form notarization records, and output row proofs, group proofs, differential proofs and constraint consistency proofs to complete the security verification and traceability management of engineering cost data.

2. The method for secure management of engineering cost data based on blockchain evidence storage according to claim 1, characterized in that, The generation of the project cost list row record set and the off-chain security library specifically includes: Obtain the project cost documents and related attachments, and categorize them according to document source, business stage, and version to form a set of documents to be processed; The set of files to be processed is subjected to format unification and structured parsing to obtain a set of normalized fields; The document metadata item set is extracted based on the normalized field set, normalized encoding and sequence solidification are performed to generate the document metadata sequence, and the summary generation operator is used to generate the summary to obtain the document metadata summary; Each list row record is solidified into a row-level structured entry based on the standardized field set, and then aggregated to generate a list of project cost list row records; The project cost documents and related attachments are encrypted and stored, generating a set of ciphertext files and a set of key encapsulations, which are then written to an off-chain security library. A set of off-chain security repository location information is generated based on the set of encrypted files. An association mapping is established between the set of off-chain security repository location information and the set of engineering cost list line records to generate an index mapping set.

3. The method for secure management of engineering cost data based on blockchain evidence storage according to claim 1, characterized in that, The generation of the row-level root specifically includes: Based on the rule version number and view label, the engineering cost list row record set is processed by field selection, field order fixing and field value normalization to obtain a row-level normalized field set; For each row-level normalized field entry in the row-level normalized field set, write the rule version number and view label in the preset position to generate a row-level normalized sequence. Then, use the digest generation operator to perform row-level hash generation to obtain row-level hash values ​​and aggregate them to generate a row-level hash set. The row-level hash set is sorted and solidified based on the unique locator key of the row-level structured entry, and then written into the leaf node sequence to generate the leaf node set; A row-level Merkle tree is constructed using the set of leaf nodes as leaf nodes. The parent node hash is generated by sequentially performing binary combinations on adjacent leaf nodes and using a digest generation operator. The nodes of each level are generated iteratively. When the number of nodes in a certain level is odd, the last node of that level is copied sequentially to make up the gap before performing binary combinations again, until a unique row-level root is generated.

4. The method for secure management of engineering cost data based on blockchain evidence storage according to claim 1, characterized in that, The generation of the grouped document root specifically includes: Obtain the row-level hash set and row-level root, and read the project ownership information, contract section ownership information and cost stage information corresponding to each row-level hash value from the project cost list row record set to form a row-level hash-cost grouping key association set; Based on the preset cost grouping key, the row-level hash-cost grouping key association set is grouped and merged to obtain the grouped row-level hash subset; For each group's row-level hash subset, perform intra-group sorting and solidification to generate a group leaf node sequence, and use these leaf nodes to construct a group Merkle tree to obtain the group root set; Establish a group root-cost group key binding relationship between the group root set and the corresponding cost group key value, and generate a group root registration table; Based on the preset cost grouping key order rules, the leaf node order of the grouping root registration table is fixed to generate the grouping root leaf node sequence, and the document-level Merkle tree is constructed as the leaf nodes to obtain the grouping document root.

5. The method for secure management of engineering cost data based on blockchain-based evidence storage according to claim 1, characterized in that, The generation of the engineering cost relationship knowledge graph specifically includes: Get the set of project cost list row records for the current version and the parent version, perform row-level alignment, and generate a version alignment mapping set; Based on the version alignment mapping set, a differential comparison is performed on the current version and the parent version of the project cost list row record set to generate a differential event list; Write each differential event in the differential event list into its corresponding unique location key, parent version attribution information, current version attribution information, and differential field set to form a structured entry set of differential events; Perform event normalization processing on the structured set of differential events to generate a normalized list of differential events; A knowledge graph of engineering cost relationships is constructed based on a normalized differential event list, an off-chain security library, an index mapping set, and a document meta-information summary, generating a knowledge graph entity set and a knowledge graph relationship set. The knowledge graph entity set and knowledge graph relation set are subjected to consistent encoding and sequential solidification to generate a knowledge graph structure sequence. The off-chain security library location information corresponding to the knowledge graph entity set in the document meta-information digest and index mapping set is written into the knowledge graph structure sequence, and the engineering cost relation knowledge graph is output.

6. The method for secure management of engineering cost data based on blockchain evidence storage according to claim 1, characterized in that, The generation of the cost grouping differential two-layer Merkle evidence storage structure specifically includes: Obtain the knowledge graph of engineering cost relationships, extract the knowledge graph entity set and knowledge graph relationship set corresponding to the differential event list, and form the differential consistency judgment input graph. Based on a preset set of hard constraints, the differential consistency judgment input graph is subjected to a constraint-satisfaction knowledge graph consistency judgment, generating a set of constraint-satisfaction results. Based on the constraint satisfaction result set, perform tree-entry gating processing on the difference event list to generate a tree-entry difference event list; Perform event-level normalization and sequence solidification on the list of possible differential events to generate a normalized differential event sequence; Construct a difference Merkle tree using the normalized difference event sequence as leaf nodes to obtain the difference root; A cost grouping differential two-layer Merkle evidence storage structure is established based on the grouped root set, grouped document root, and differential root.

7. The method for secure management of engineering cost data based on blockchain-based evidence storage according to claim 1, characterized in that, The generation of the aforementioned engineering cost data security verification and traceability management specifically includes: Obtain the cost grouping differential two-layer Merkle evidence storage structure, document metadata summary, constraint satisfaction result set, and multi-subject signature set involved in engineering cost management; generate constraint satisfaction summary based on constraint satisfaction result set, and generate signature set summary based on multi-subject signature set; The constraint satisfaction summary, signature set summary, document metadata summary, and cost grouping differential two-layer Merkle evidence storage structure are solidified to generate an anchored input sequence. An anchor hash is generated by using a digest generation operator on the anchor input sequence to obtain the on-chain anchor. An on-chain anchor-index mapping and binding relationship is established between the on-chain anchor and the off-chain security library location information set to generate an on-chain anchor registration table. The on-chain anchor, cost group differential two-layer Merkle notarization structure, document meta-information digest, constraint satisfaction digest, signature set digest, and on-chain anchor registration table are written into the consortium blockchain to form notarization records, and a notarization record index information set is generated. Row proofs are generated based on row-level Merkle trees, group proofs are generated based on grouped Merkle trees, difference proofs are generated based on difference Merkle trees, and constraint consistency proofs are generated based on the constraint satisfaction result set and the difference consistency determination input graph. During the verification and traceability phase, the system receives the line records or differential events of the engineering cost list to be verified, locates the evidence record index information set based on the on-chain anchor registration table, and reads the cost group differential two-layer Merkle evidence storage structure, document meta information digest, constraint satisfaction digest, and signature set digest corresponding to the on-chain anchor. The system then performs path recalculation and consistency verification using line proof, group proof, differential proof, and constraint consistency proof, outputs the verification conclusion, and completes the traceability management of engineering cost data when the consistency verification passes.

Images