Intelligent project financing decision support system based on full life cycle

Abstract

The invention relates to the technical field of financing decision, in particular to an intelligent project financing decision support system based on a full life cycle, which comprises the steps of acquiring task information and constructing a task fund path, identifying a node triggering and fund corresponding relation, analyzing delay influence to form a risk propagation path, and rearranging key node time to generate a release sequence table. And measuring fund release rhythm and periodic pressure, and summarizing task paths and fund order to generate a full-period financing decision basis set. The method comprises the following steps of: constructing a task and fund path link; establishing a node triggering and fund release corresponding structure; identifying a risk conduction chain of a delay task and extracting an interference path; rearranging key node time to generate a release priority sequence; a whole-process task fund connection structure is formed, and fund scheduling coherence and risk response matching degree are enhanced.

Description

Technical Field

[0001] This invention relates to the field of financing decision-making technology, and in particular to an intelligent project financing decision support system based on the entire project lifecycle. Background Technology

[0002] The field of financing decision-making technology encompasses financial analysis, investment management, risk control, and project evaluation. Its core content involves quantitatively assessing and judging key factors such as investment direction, scale, and return cycle based on economic data, market information, and corporate operations. This assists managers or decision-makers in making rational and systematic capital allocations. It mainly covers multiple stages including data collection, economic modeling, financial forecasting, risk assessment, and decision rule setting, and is widely used in banking, insurance, securities, funds, and corporate investment and financing scenarios. With the development of information processing methods such as big data, artificial intelligence, and information systems, financing decision-making is gradually moving towards intelligence, automation, and end-to-end integration, placing higher demands on information accuracy, model adaptability, and decision-making timeliness.

[0003] Among them, the intelligent project financing decision support system based on the entire life cycle refers to a decision support system that systematically provides suggestions for financing scheme selection and dynamic adjustment throughout the entire project implementation process, from project initiation, feasibility study, investment decision-making, execution monitoring to post-evaluation. It primarily addresses technical matters such as identifying funding needs, assessing risks, and determining funding matching mechanisms at different stages of project financing, making comprehensive judgments based on economic benefit evaluation methods, cash flow analysis models, and debt repayment capacity assessment methods. Its support includes using time-series financial data analysis to calculate investment returns, applying static and dynamic matching methods for capital structure design, and combining a multi-dimensional indicator system for credit assessment and investment and financing model optimization to provide financing path solutions with higher matching accuracy.

[0004] Existing technologies lack the ability to identify the path structure of the connection between task execution and fund release. When the node triggering and fund release distribution are inconsistent, it is easy to cause sequence misalignment. It is difficult to track the risk chain after the delay event occurs and its spread in the task sequence. The scope of impact cannot be accurately quantified. The adjustment of the fund release order cannot be dynamically optimized according to the specific interference structure. There is a lack of linkage identification means between the fund release rhythm and task intensity within the cycle. It is easy for task overlap and fund allocation imbalance to occur during the concentrated fund release phase, resulting in excessive pressure in local periods and forming phased liquidity risks. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by proposing an intelligent project financing decision support system based on the entire project lifecycle.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: an intelligent project financing decision support system based on the entire project lifecycle, the system comprising: The task link parsing module obtains all task information of the project, identifies the time connection sequence between each task, locates the corresponding events of fund release and task nodes, marks the connection structure between all task nodes and fund behavior, draws the path clue diagram between tasks and funds in sequence, and generates a task and disbursement connection structure diagram. The risk transmission identification module reads the task path in the task and disbursement connection structure diagram, determines whether the sequential relationship between tasks causes a change in the order before and after fund release, identifies the segment where continuous tasks form release interference, records the distribution of interference paths and node sequence positions, and generates a risk transmission distribution map. The fund release sorting module extracts the original time arrangement order based on the high interference nodes identified in the risk transmission distribution map, rearranges the corresponding release time order, updates the node ranking and adjusts the release sorting, and generates a risk priority release order table. The periodic pressure measurement module calls the time period information in the risk priority release sequence table, identifies the changes in task concentration and release distribution during the fund release time period, screens dense release segments, marks time segments with continuous changes in fluctuation amplitude, and generates a periodic fund pressure distribution map.

[0007] As a further aspect of the present invention, the task and disbursement linkage structure diagram includes a task node sequence structure, fund release event nodes, task and fund pairing relationship, and path connection mapping elements; the risk transmission distribution diagram includes a set of delay impact paths, a time continuous interference segment identifier, and a fund release association chain; the risk priority release order table specifically includes a list of key fund nodes, a time order adjustment identifier, and a risk impact priority sequence; and the periodic fund pressure distribution diagram includes concentrated release rhythm segments, dense task arrangement intervals, and fund release time pressure points.

[0008] As a further aspect of the present invention, the task link parsing module includes: The task data extraction submodule obtains the task node number, task time tag, and task description field from the task configuration list during the project execution phase. It checks whether the planned start time and end time fields corresponding to each task are complete, calls the task node number and time tag to construct a task time series array, arranges each task number in sequence according to the task time order and marks its corresponding time interval, and generates a task time series structure matrix. The task trigger structure identification submodule determines whether there are overlapping or gap intervals between adjacent task time intervals based on the task number order in the task time sequence structure matrix. If the start time of a task is earlier than the end time of the previous task, it is marked as an overlapping trigger type. If the start time is equal to the end time of the previous task, it is marked as a continuous trigger type. If there is a gap, it is marked as a delayed trigger type. The marking results and task node numbers are used to construct a trigger link logic array and generate a trigger type identification vector between tasks. The task allocation path generation submodule, based on the task number order in the inter-task trigger type identifier vector, calls the fund behavior number associated with the corresponding task node in the fund allocation record table, matches whether the fund allocation time and the task execution time overlap. If the time overlaps, a connection structure is established between the fund behavior number and the task number, constructs a one-to-one mapping matrix between task nodes and fund behavior nodes, draws a path clue diagram composed of fund behavior nodes in the order of task nodes, and generates a task allocation connection structure diagram.

[0009] As a further aspect of the present invention, the process of generating the task time series structure matrix specifically includes: creating the task time series structure matrix, which includes a task node number column, a planned start time column, and an end time column; filling the task node number, the planned start time, and the end time into the corresponding columns of the task time series structure matrix, and sorting the task time series structure matrix in ascending order according to the planned start time; The process of matching whether the fund disbursement time and the task execution time coincide is as follows: obtain the fund disbursement time corresponding to the fund behavior number in the fund disbursement record table; extract the planned start time and the end time corresponding to the task number from the task time sequence structure matrix; determine whether the fund disbursement time is within the closed time interval formed by the planned start time and the end time; if the fund disbursement time is within the closed time interval, it is determined that the fund disbursement time and the task execution time coincide.

[0010] As a further aspect of the present invention, the risk transmission identification module includes: The path structure parsing submodule, based on the task path structure in the task and disbursement connection structure graph, obtains the task node number and its corresponding fund release time in each path, detects whether the time interval between adjacent task nodes meets the continuation triggering condition, determines whether the time offset of the preceding node causes the change in the start time of the subsequent node, filters out task combinations with preceding and following relationships, and generates a list of delay impact transmission paths. The interference segment determination submodule calculates the cumulative time offset of each node on the task path based on the time difference of the task pairs in the delay effect transmission path list, compares it with the set continuous interference threshold of task time, determines whether there is a continuous node segment that meets the interference interval length requirement, calls the cumulative time offset sequence and the task node order, marks the formed continuous interference node set, and obtains the continuous interference segment interval vector. The transmission trajectory construction submodule calls the task number combination in the interval vector of the continuous interference segment, matches its associated fund release number and release time point, connects the fund release nodes with continuous task interference relationship, arranges the fund release behavior according to the time sequence, draws the impact extension trajectory on each path, and establishes a risk transmission distribution map.

[0011] As a further aspect of the present invention, the fund release sorting module includes: The key node extraction submodule, based on the path structure in the risk transmission distribution map, detects whether the task node corresponding to the fund release node in the path is within the marked range of the continuous interference segment interval vector, extracts the set of task numbers that meet the continuous interference condition, obtains their original fund release time label, and forms a mapping relationship with the task number to generate an interference key node time index table. The release priority adjustment submodule calls the original time tag sequence of the fund release nodes in the time index table of the interference key nodes, sorts them in ascending order according to the time sequence, and reassigns a new priority identifier to each fund release node. At the same time, it constructs a difference matrix between the old priority and the new priority, establishes a release priority adjustment structure, and obtains a fund release priority rearrangement matrix. The sorting structure generation submodule recombines the fund release number with its adjusted order index according to the new order identifier in the fund release order rearrangement matrix, generates a complete list of release numbers and a corresponding list of release times, constructs an updated time sorting mapping table corresponding to the risk path structure, and establishes a risk priority release order table.

[0012] As a further aspect of the present invention, the periodic pressure measurement module includes: The release rhythm identification submodule, based on the fund release time field in the risk priority release sequence table, counts the number of fund release events in each time period, calculates the release frequency difference between adjacent time periods, determines whether the release frequency change exceeds the set release frequency fluctuation threshold, filters out time segments that continuously exceed the fluctuation threshold, and generates a list of concentrated fund release intervals. The dense section determination submodule calls the time period information in the list of concentrated fund release intervals to obtain the task node number and its arrangement time interval within the corresponding period. It calculates the task arrangement density coefficient based on the overlap of the time intervals, compares the arrangement density coefficient with the number of fund release events in parallel, and filters out the period positions where tasks are dense and release behavior occurs continuously to obtain the high-pressure release period segment identification table. The cycle pressure plotting submodule extracts the corresponding time point and quantity value of fund release events based on the cycle number in the high-pressure release cycle segment identifier table, constructs the release frequency sequence for each cycle unit, arranges the frequency sequence in ascending order by cycle number, matches its position interval in the total cycle, plots a visualization graphic of the release intensity under the corresponding cycle number, and establishes a cycle fund pressure distribution map.

[0013] As a further aspect of the present invention, the process of counting the number of fund release events in each time period specifically involves: setting the statistical period unit to a natural day; traversing the fund release time field in the risk priority release order table, and counting the number of fund release events falling within the same natural day. The process of calculating the task arrangement density coefficient based on the overlap of time intervals is as follows: obtain all the task node numbers corresponding to the time period in the list of concentrated fund release intervals; extract the arrangement time intervals corresponding to all the task node numbers, calculate the total duration of all the arrangement time intervals within the time period, and calculate the overlap duration of any two arrangement time intervals within the time period; sum the overlap durations, and divide the sum of the overlap durations by the total duration to obtain the task arrangement density coefficient.

[0014] As a further aspect of the present invention, the system further includes: The full-cycle decision generation module reads the pressure release segment in the cycle capital pressure distribution map, extracts the relationship between task distribution and release rhythm, identifies the task order rearrangement and combination, organizes the corresponding structure of task segments and capital release time lines, and generates a full-cycle financing decision basis set. The full-cycle financing decision-making basis set includes stage combination classification items, key task time sequence position, and task funding path overview table.

[0015] As a further aspect of the present invention, the full-cycle decision generation module includes: The phase combination extraction submodule, based on the period segment number in the periodic capital pressure distribution map, detects whether there is an overlapping interval between the peak of capital release frequency and the task arrangement, extracts the period number that meets the dual conditions of concentrated release and overlapping tasks, records the associated task number and capital number under each period, establishes the correspondence between period and task capital combination, and generates a concentrated phase combination list. The instruction node integration submodule calls the set of task numbers and fund numbers in the centralized stage combination list, obtains the time tag of the task and the corresponding fund release priority identifier, groups the fund release behavior within the stage according to the start and end time of the task, integrates the path sequence of fund release behavior in chronological order, constructs the combination structure of task segment and fund behavior segment, and obtains the stage task fund path mapping table. The decision basis summary submodule identifies the sequence of fund release nodes in each stage based on the task fund segment combination structure in the stage task fund path mapping table, standardizes it into a standard instruction format, summarizes the task number, fund number and sequence number fields of all stage combinations, integrates them into a continuous executable data structure sequence, and establishes a full-cycle financing decision basis set.

[0016] Compared with the prior art, the advantages and positive effects of the present invention are as follows: 1. In this invention, by constructing a path clue structure between task nodes and fund release behavior, the time sequence labeling and fund event mapping between preceding and subsequent nodes are completed, the correspondence between task triggering order and fund flow is sorted out, and a structural linkage of fund release during task advancement is established. This supports the organization of fund behavior and task chain matching under multiple nodes and multiple paths, forming a connection basis for fund allocation between stages.

[0017] 2. In this invention, the interference range is identified based on the delay segments in the task path, nodes with continuous time impact are extracted, a chain structure for the propagation of risk along the task path is constructed, and a sequence of fund release order is formed by combining the time position of key nodes, thereby reconstructing the order of fund release. The release rhythm is adjusted based on the interference path to maintain structural consistency between fund flow and risk pressure, and to reduce the amplification effect in local links.

[0018] 3. In this invention, the rhythm changes and task arrangement density during the fund release process are compared and analyzed, the concentrated release segments are extracted and the periodic pressure positions are marked, the task paths and fund release order are aggregated according to the stages, and a task fund sequence arrangement covering the entire process is constructed to enhance the coordination and matching between different time periods, alleviate the periodic conflicts caused by concentrated resource investment, and improve the temporal balance of fund supply and the sustainability of resource allocation. Attached Figure Description

[0019] Figure 1 This is an overall flowchart of the system of the present invention; Figure 2 This is a flowchart of the modules of the system of the present invention. Detailed Implementation

[0020] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0021] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0022] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0023] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0024] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0025] Please see Figure 1 This invention provides a technical solution: an intelligent project financing decision support system based on the entire project lifecycle, the system comprising: The task link parsing module obtains all task information during the project execution phase, identifies the temporal connection sequence between each task, confirms the preceding and following nodes on which the triggering relationship depends, locates the corresponding events of fund release and task nodes, marks the connection structure between all task nodes and fund behavior, draws the path clue diagram between tasks and funds in sequence, and generates a task and disbursement connection structure diagram. The risk transmission identification module reads the task path in the task and disbursement connection structure diagram, identifies the extension effect of the delay of the preceding node on the subsequent node, locates the transmission relationship between the fund release nodes, determines whether the impact forms continuous interference on the timeline, connects the continuous impact segments to form a path set, draws the trajectory diagram of the fund release impact propagation on each path, and generates a risk transmission distribution map. The fund release sorting module identifies key nodes with continuous interference segments in the path based on the path structure in the risk transmission distribution diagram, extracts the original time arrangement positions of these nodes, rearranges the time identifiers corresponding to the fund release order, adjusts the original release order order, forms a time sequence update table corresponding to the risk structure, and generates a risk priority release order table. The cycle pressure measurement module calls the time arrangement in the risk priority release sequence table to identify whether there is a concentrated tilt in the pace of fund release, extracts the task information of the concentrated release time period, compares the task arrangement density of adjacent stages, marks the cycle segments that are continuously changing in the dense arrangement, draws the cycle position map corresponding to the dense release behavior, and generates a cycle fund pressure distribution map. The full-cycle decision generation module reads the cycle segments in the cycle funding pressure distribution map, identifies the combination of stages with concentrated releases and overlapping tasks, confirms the position changes of key tasks in the combination, integrates the release time and task relationship to form a stage classification, summarizes the task funding release path and node order of all segments, and forms a unified full-stage instruction arrangement set to generate a full-cycle financing decision basis set.

[0026] The task and disbursement linkage structure diagram includes the task node sequence structure, fund release event nodes, task and fund pairing relationship, and path connection mapping elements. The risk transmission distribution diagram includes the set of delay impact paths, the identification of continuous time interference segments, and the fund release association chain. The risk priority release order table specifically includes a list of key fund nodes, time sequence adjustment indicators, and risk impact priority sequence. The cycle fund pressure distribution diagram includes concentrated release rhythm segments, dense task arrangement intervals, and fund release time pressure points. The full cycle financing decision basis set includes stage combination classification items, key task time sequence positions, and a task fund path overview table.

[0027] Please see Figure 2 The task link parsing module includes: The task data extraction submodule obtains the task node number, task time tag, and task description field from the task configuration list during the project execution phase. It checks whether the planned start time and end time fields corresponding to each task are complete, calls the task node number and time tag to construct a task time series array, arranges each task number in sequence according to the task time order and marks its corresponding time interval, and generates a task time series structure matrix. Retrieve the "Project Task Configuration Table" from the project management database. This table records task node numbers, task time tags, and task description fields, as shown in Table 1. For each row of task data extracted from this table, check whether its "Planned Start Time" and "End Time" fields contain valid time unit values. For example, check the planned start time "Day 1" and end time "Day 10" for task T-001. Both are valid positive integers, so the data is considered complete. Continue checking task T-002; its start time "Day 10" and end time "Day 20" are both valid, so the data is considered complete. Task T-003 has a valid start time "day 18" and end time "day 25", so it is considered complete. When task T-004 is detected, its "end time" field is found to be missing, so it is considered incomplete. Therefore, this data T-004 will be marked and excluded from subsequent processing. Task T-005 is detected, and its valid start time "day 28" and end time "day 35" are considered complete. Subsequently, the task node numbers and timestamps of all tasks determined to have complete data are retrieved to construct a temporary task time series array. Its data structure is [(T-001, Based on the array [T-001, 1, 10), (T-002, 10, 20), (T-003, 18, 25), (T-005, 28, 35)], a task time series structure matrix is ​​created. This matrix is ​​set to contain three columns: "task node number", "planned start time", and "end time". The data items in the array, such as (T-001, 1, 10), are filled into the corresponding three columns in the first row of the matrix. In this way, the data for T-002, T-003, and T-005 are filled in. Finally, all rows of the matrix are sorted in ascending order strictly according to the value of the "planned start time" column. Since the sequence [1, 10, 18, 28] is already ordered, the row order of the sorted matrix remains unchanged as T-001, T-002, T-003, T-005, thus generating the task time series structure matrix.

[0028] Table 1: Project Task Configuration Table ; As shown in Table 1, this table is a general project task configuration example. Task T-004 was removed in the data integrity check because the end time field was missing.

[0029] The task trigger structure identification submodule determines whether there are overlapping or gap intervals between adjacent task time intervals based on the task number order in the task time sequence structure matrix. If the start time of a task is earlier than the end time of the previous task, it is marked as an overlapping trigger type. If the start time is equal to the end time of the previous task, it is marked as a continuous trigger type. If there is a gap, it is marked as a delayed trigger type. The marking results and task node numbers are used to construct a trigger link logic array and generate a trigger type identification vector between tasks. Based on the task numbering order in the task time series structure matrix, i.e., [T-001, T-002, T-003, T-005], starting from the second row of the matrix (i.e., task T-002), we sequentially determine whether adjacent task time intervals overlap, are consecutive, or have gaps. First, we extract the planned start time (day 10) of T-002 and the planned end time (day 10) of T-001. We then perform a judgment: comparing the start time 10 of T-002 with the end time 10 of T-001, i.e., 10 equals 10, the condition is met, so the relationship between T-001 and T-002 is marked as "consecutive trigger type". Next, we process T-003 and T-002, extracting the planned start time (day 18) of T-003 and the planned end time (day 20) of T-002. We then perform a judgment: comparing the start time 18 of T-003 with the end time 20 of T-002, and determine... If 18 is earlier than 20, the condition is met, so the relationship between T-002 and T-003 is marked as "overlapping trigger type". Finally, T-005 and T-003 are processed. The planned start time of T-005 (day 28) and the planned end time of T-003 (day 25) are extracted. The following judgments are made: Is 28 equal to 25? The condition is not met. Then, is 28 earlier than 25? The condition is not met. The difference between the two time units is calculated, i.e., 28 minus 25, which gives a gap of 3 days. This value is greater than 0, so a gap is determined to exist. Therefore, the relationship between T-003 and T-005 is marked as "delayed trigger type". After completing the judgment of all adjacent task pairs, the marking results and the corresponding task node numbers are called to construct a trigger link logic array. The content of this array is [(T-001, [T-002, Continuous Trigger), (T-002, T-003, Overlapping Trigger), (T-003, T-005, Delayed Trigger)], after being formatted, this array generates a vector of inter-task trigger type identifiers.

[0030] The task allocation path generation submodule, based on the task number order in the inter-task trigger type identifier vector, calls the fund behavior number associated with the corresponding task node in the fund allocation record table, matches whether the fund allocation time and the task execution time overlap. If the time overlaps, it establishes a connection structure between the fund behavior number and the task number, constructs a one-to-one mapping matrix between task nodes and fund behavior nodes, draws a path clue diagram composed of fund behavior nodes in the order of task nodes, and generates a task allocation connection structure diagram. Based on the task number sequence T-001, T-002, T-003, T-005 determined by the inter-task trigger type identifier vector, the "Funds Allocation Record Table" is queried. The specific contents of this table are shown in Table 2. The fund behavior number associated with each task node is called. For example, T-001 is associated with F-01, T-002 is associated with F-02, T-003 is associated with F-03, and T-005 is associated with F-04. Then, the fund allocation time is matched one by one to see if it coincides with the task execution time. Specifically, the fund allocation time "day 5" of F-01 is obtained, and the planned start time "day 1" and end time "day 10" of T-001 are extracted from the task time sequence structure matrix to construct a closed time interval [1, 10], Execute the judgment: Determine whether the payment time 5 is greater than or equal to the start time 1 of the interval, and whether 5 is less than or equal to the end time 10 of the interval. Both conditions are met, so it is determined that F-01 and T-001 have overlapping times. Then process T-002, whose interval is [10, 20]. The payment time of F-02 is "day 15". Execute the judgment: 15 is greater than or equal to 10 and 15 is less than or equal to 20. The condition is met, so it is determined that the times overlap. Then process T-003, whose interval is [18, 25]. The payment time of F-03 is "day 26". Execute the judgment: 26 is greater than or equal to 18 but 26 is not less than or equal to 25. The condition is not met, so it is determined that the times do not overlap. Then process T-005, whose interval is [28,

[35] The payment time for F-04 is "day 30". The judgment is: 30 is greater than or equal to 28 and 30 is less than or equal to 35. If the condition is met, it is determined that the time coincides. If the time coincides, a connection structure is established between the fund behavior number and the task number. All pairs determined to coincide in time [(T-001, F-01), (T-002, F-02), (T-005, F-04)] are stored in a one-to-one mapping matrix. When drawing the path of the fund behavior node, according to the task node order (T-001 -> T-002 -> T-003 -> T-005), only the nodes that satisfy the mapping relationship are connected, that is, F-01 -> F-02. Since T-003 did not match successfully, its connection is interrupted. F-04 corresponding to T-005 forms a separate path node and generates a task payment connection structure diagram.

[0031] Table 2: Project Fund Disbursement Record Sheet ; See Table 2, which provides information on the relationship between fund disbursement and task nodes. The disbursement time of F-03 (day 26) exceeds the planned interval [18, 25] of its associated task T-003.

[0032] The risk transmission identification module includes: The path structure parsing submodule, based on the task path structure in the task and disbursement connection structure graph, obtains the task node number and its corresponding fund release time in each path, detects whether the time interval between adjacent task nodes meets the continuation triggering condition, determines whether the time offset of the preceding node causes the change of the start time of the subsequent node, filters out task combinations with preceding and following relationships, and generates a list of delay impact transmission paths. Based on the task path structure in the task and disbursement linkage diagram, especially the main path T-001 -> T-002 -> T-003 (although T-003 has mismatched funding, the task flow still exists), and by incorporating "Project Actual Execution Log" data, the task node numbers T-001, T-002, and T-003 within this path are obtained. T-003, and their actual execution times recorded in the log, assuming the log records: T-001's actual end time is "day 12" (planned day 10), T-002's actual start time is "day 13" (planned day 10), and its actual end time is "day 25" (planned day 20), and T-003's actual start time is "day 26" (planned day 18), and its actual end time is "day 36" (planned day 25). First, the time interval between T-001 and T-002 is checked. T-002's actual start time (day 13) is later than T-001's actual end time (day 12), satisfying the continuation triggering condition. Next, determine whether the time offset of the preceding node T-001 causes a change in the start time of the following node T-002. Calculate the time offset of T-001, i.e., (T-001 actual end time "day 12") minus (T-001 planned end time "day 10"), resulting in an offset of 2 days. Calculate the change in the start time of T-002, i.e., (T-002 actual start time "day 13") minus (T-002 planned start time "day 10"), resulting in a change of 3 days. Since the change in the start time of T-002 for 3 days is not zero, it is determined that the offset of T-001 has affected the start time of T-002. Therefore, (T-001, As a task combination with a sequential relationship, (T-002) is further analyzed. The actual start time of T-003 (day 26) is later than the actual end time of T-002 (day 25), satisfying the continuation trigger. The time offset of T-002 (25 days minus 10 days) is calculated to be 5 days, and the change in the start time of T-003 (26 days minus 18 days) is calculated to be 8 days. Since 8 days is not 0, (T-002, T-003) is also determined to be a related combination. All the selected combinations [(T-001, T-002), (T-002, T-003)] are gathered to generate a list of delay impact transmission paths.

[0033] The interference segment determination submodule calculates the cumulative time offset of each node on the task path based on the time difference of the task pairs in the delay effect propagation path list, and compares it with the set continuous interference threshold of task time to determine whether there is a continuous node segment that meets the interference interval length requirement. It calls the cumulative time offset sequence and the task node order to mark the set of continuous interference nodes and obtain the continuous interference segment interval vector. Based on the list of delay impact propagation paths [(T-001, T-002), (T-002, T-003)], calculate the cumulative time offset for each node along this path. Here, the cumulative time offset is defined as the difference between the actual end time and the planned end time of a node. For node T-001, the cumulative time offset is (T-001 actual end time "day 12") minus (T-001 planned end time "day 10"), resulting in 2 days. For node T-002, the cumulative time offset is (T-002 actual end time "day 25") minus (T-002 planned end time "day 20"), resulting in 5 days. For node T-003, the cumulative time offset is (T-003 actual end time "day 36") minus (T-003 planned end time "day 25"), resulting in 11 days. The resulting offset sequence is [2, 5,

[11] Next, this sequence is compared with the set continuous interference threshold for task time. The threshold is set with reference to the analysis of retrospective data of 50 completed projects of the same type. Statistical analysis shows that when the total cumulative delay time of continuous tasks exceeds 10 time units (days), the probability of project failure or major reorganization increases to 85%. Therefore, based on this statistical result, the continuous interference threshold for task time is set to 10 days. The offset sequence is compared one by one: the offset of T-001 is 2 days, which is less than 10 days; the offset of T-002 is 5 days, which is less than 10 days; and the offset of T-003 is 11 days, which is greater than or equal to 10 days. At this time, node T-003 is determined to meet the interference condition, and all its predecessor nodes T-001 and T-002 on the transmission path are traced back. {T-001, T-002, T-003} are marked as the continuous interference node set, and the continuous interference segment interval vector is obtained.

[0034] The transmission trajectory construction submodule calls the task number combination in the interval vector of the continuous interference segment, matches its associated fund release number and release time point, connects the fund release nodes with continuous task interference relationship, arranges the fund release behavior according to the time sequence, draws the impact extension trajectory on each path, and establishes a risk transmission distribution map. The obtained continuous interference segment interval vector {T-001, T-002, T-003} is called, and the task number combination is extracted. This combination is then matched with the associated fund behavior number in the fund disbursement record table (Table 2). T-001 is associated with F-01, T-002 with F-02, and T-003 with F-03 (even though F-03's time does not overlap in segment 3, it is still a task-associated fund node of T-003). Simultaneously, the corresponding fund release time points are obtained: "Day 5" for F-01, "Day 15" for F-02, and "Day 26" for F-03. Next, the fund release nodes with continuous task interference relationships are connected. Since tasks T-001, T-002, and T-003 are identified as continuous interference at the task level, their corresponding fund nodes F-01 and F-02 are... F-03 is connected by directed line segments, and these fund release actions are arranged in chronological order according to the time of fund release, namely F-01 (day 5) -> F-02 (day 15) -> F-03 (day 26). This sequence constitutes the trajectory of the impact at the fund level. When drawing the graph, the task offset of T-001 is marked as "2 days" on the line connecting F-01 and F-02, the task offset of T-002 is marked as "5 days" on the line connecting F-02 and F-03, and the offset of T-003 is marked as "11 days" on the F-03 node itself. In this way, the delay risk at the task level is mapped to the time series of fund disbursement, and a risk transmission distribution map is established.

[0035] The fund release sorting module includes: The key node extraction submodule, based on the path structure in the risk transmission distribution map, detects whether the task node corresponding to the fund release node in the path is within the marked range of the continuous interference segment interval vector, extracts the set of task numbers that meet the continuous interference condition, obtains their original fund release time label, and forms a mapping relationship with the task number to generate an interference key node time index table. Based on the path structure F-01 -> F-02 -> F-03 in the risk transmission distribution map, the task nodes T-001, T-002, and T-003 corresponding to each fund release node F-01, F-02, and F-03 within the path are detected. It is determined whether they fall within the marked range of the continuous interference segment interval vector {T-001, T-002, T-003} determined in the previous steps. Specifically, for F-01, its associated task node T-001 in Table 2 is called, and T-001 is compared with the vector {T-001, T-002, T-003}. The membership of task IDs {T-001, T-002, T-003} is checked. T-001 exists in the set, so the check result is true. Next, for F-02, its associated task T-002 is called to check if T-002 exists in the set, and the check result is true. Finally, for F-03, its associated task T-003 is called to check if T-003 exists in the set, and the check result is true. All nodes meet the conditions. Therefore, the task ID set {T-001, T-002, T-003} that satisfies the continuous interference condition is extracted. Subsequently, the original fund release time tags corresponding to these task IDs are obtained from the fund disbursement record table (Table 2). These are: T-001 corresponds to the time "day 5" of F-01, T-002 corresponds to the time "day 15" of F-02, and T-003 corresponds to the time "day 26" of F-03. A one-to-one mapping relationship is established between the task ID, fund ID, and their fund release time tags, constructing a set containing (task ID, fund ID, ...). The data structure for the release time is [(T-001, F-01, 5), (T-002, F-02, 15), (T-003, F-03, 26)], which generates a time index table for key interference nodes.

[0036] The release priority adjustment submodule calls the original time tag sequence of fund release nodes in the time index table of interference key nodes, sorts them in ascending order according to time sequence, and reassigns a new priority identifier to each fund release node. At the same time, it constructs the difference matrix between the old priority and the new priority, establishes the release priority adjustment structure, and obtains the fund release priority rearrangement matrix. Retrieve the original time stamp sequence of the fund release nodes from the time index table of key interference nodes, i.e., [day 5, day 15, day 26], and obtain their corresponding fund numbers [F-01, F-02, F-03]. Sort the time stamp sequence in ascending order according to time sequence. The sorted sequence is [5, 15, ...].

[26] Remaining unchanged, based on the sorting results, each fund release node is reassigned a new priority identifier. F-01 (day 5) is assigned a new priority of 1, F-02 (day 15) is assigned a new priority of 2, and F-03 (day 26) is assigned a new priority of 3. Simultaneously, the old priority identifiers of these fund nodes are retrieved from the "Initial Project Disbursement Plan." Assuming F-01's old priority is 1, F-03's old priority is 2, and F-02's old priority is 3 (i.e., F-03 was originally scheduled to be disbursed before F-02), a difference matrix between the old and new priorities is constructed. This matrix contains four columns: Fund Number, Old Priority, New Priority, and Difference Value. The difference value (new priority minus old priority) is calculated row by row: For F-01, the difference is 1 minus 1, equal to 0; for F-02, the difference is 2 minus 3, equal to -1; and for F-03, the difference is 3. Subtracting 2 equals +1, resulting in the matrix [[F-01, 1, 1, 0], [F-02, 3, 2, -1], [F-03, 2, 3, +1]]. This matrix is ​​the release priority adjustment structure, yielding the fund release order rearrangement matrix.

[0037] The sorting structure generation submodule recombines the fund release number with its adjusted order index according to the new order identifier in the fund release order rearrangement matrix, generates a complete list of release numbers and a corresponding list of release times, constructs an updated time sorting mapping table corresponding to the risk path structure, and establishes a risk priority release order table. Based on the new order identifiers in the fund release order rearrangement matrix (F-01 corresponds to 1, F-02 to 2, and F-03 to 3), the fund release numbers and their adjusted order indices are recombined. Specifically, the fund release number column [F-01, F-02, F-03] and the new order index column [1, 2, 3] are extracted and combined into a new list. Each element is a tuple containing the release number and the new order, such as [(F-01, 1), (F-02, 2), (F-03, 3)]. A complete release number list [F-01, F-02, F-03] and a corresponding release time list [Day 5, Day 15, Day 26] extracted from the interference key node time index table are generated based on this list. The indices of these two lists correspond one-to-one. Subsequently, a risk path structure (T-001 -> T-002 -> ...) is constructed. The updated time sorting mapping table corresponding to T-003 integrates task, funds, new priority, and time. The table contains four fields: "Task Number", "Fund Number", "New Priority", and "Release Time". Data is filled in row by row. The first row is (T-001, F-01, 1, 5), the second row is (T-002, F-02, 2, 15), and the third row is (T-003, F-03, 3, 26). A risk priority release order table is established.

[0038] The periodic pressure measurement module includes: The release rhythm identification submodule, based on the fund release time field in the risk priority release sequence table, counts the number of fund release events in each time period, calculates the release frequency difference between adjacent time periods, determines whether the release frequency change exceeds the set release frequency fluctuation threshold, filters out time segments that continuously exceed the fluctuation threshold, and generates a list of concentrated fund release intervals. Based on the risk priority release sequence table and all other fund disbursement records in the project, all fund release time fields are extracted. [F-01(5), F-02(15), F-03(26), F-04(30)] are obtained from Table 2. It is assumed that there are other non-risk path disbursements F-05 (day 15) and F-06 (day 16). The complete set of release time points is [5, 15, 15, 16, 26, 30]. The number of fund release events within each time period is counted. Specifically, the statistical period unit is set to "natural day". All release time points are iterated through, and the number of events falling on the same natural day is counted to obtain the daily release frequency sequence (taking day 14 to day 17 as an example): (day 14, 0 times), (day 15, 2 times [F-02, F-05]), (day 16, 1 time [F-06]), (day 17, Next, calculate the difference in release frequency between adjacent time periods (days). Calculate the difference between day 15 and day 14: |2 minus 0| equals 2. Calculate the difference between day 16 and day 15: |1 minus 2| equals 1. Calculate the difference between day 17 and day 16: |0 minus 1| equals 1. Determine whether the change in release frequency exceeds the set release frequency fluctuation threshold. This threshold is set according to the liquidity management regulations of the project finance department. The regulations stipulate that a warning is triggered if the number of daily payment events changes by more than 1.5 (i.e., changes by 2 times or more). Therefore, the threshold is set to 1.5. Compare the calculated differences: difference 2 is greater than 1.5, difference 1 is less than 1.5, difference 1 is less than 1.5. Filter out the continuous time intervals where the difference exceeds the threshold, that is, the interval formed by the sudden increase from 0 times on day 14 to 2 times on day 15 [day 14, day 15], and generate a list of concentrated fund release intervals.

[0039] The dense section determination submodule calls the time period information in the list of concentrated fund release intervals, obtains the task node number and its arrangement time interval within the corresponding period, calculates the task arrangement density coefficient based on the overlap of the time interval, compares the arrangement density coefficient with the number of fund release events in parallel, and filters out the period positions where tasks are dense and release behavior occurs continuously, thus obtaining the high-pressure release period segment identification table. The time period information from the fund release concentrated interval list, i.e., [day 14, day 15], is retrieved. All task node numbers executed within the corresponding period (day 14 or day 15) are obtained. Table 1 and supplementary task data are queried. T-002 [10, 20] is active within this interval. It is assumed that two other tasks, T-006 [day 12, day 16] and T-007 [day 14, day 17], are also active within this interval. The time intervals for these tasks are extracted, and the task density coefficient within this time period [day 14, day 15] (total analysis duration is 2 days) is calculated. Specifically, the total duration of all intervals within this time period is calculated. T-002's duration within [14, 15] is 2 days, T-006's duration within [14, 15] is 2 days, and T-007's duration within [14, 15] is 2 days, for a total duration of 2 + 2 + 2 = 6 days. Calculate the overlap duration of any two time intervals within this time period. The overlap duration of T-002 and T-006 in [14, 15] is 2 days, the overlap duration of T-002 and T-007 in [14, 15] is 2 days, and the overlap duration of T-006 and T-007 in [14, 15] is 2 days. Summing the cumulative overlap durations, we get 2 + 2 + 2 = For 6 days, the cumulative sum of overlapping time (6 days) is divided by the total time (6 days) to obtain a task arrangement density coefficient of 1.0. Then, the arrangement density coefficient of 1.0 is compared in parallel with the number of fund release events (the peak of this interval, i.e., 2 times on day 15). The task density threshold is set to 0.7 (based on historical project data, a coefficient greater than 0.7 indicates high task concurrency), and the release behavior duration threshold is set to 1.5 (i.e., the number of releases per day is greater than 1.5, which is 2 times in actual operation). The judgment is executed: (density coefficient 1.0 > threshold 0.7) and (release quantity 2 > threshold 1.5). Both conditions are met, so the cycle position of [day 14, day 15] is selected to obtain the high-pressure release cycle segment identification table.

[0040] The cycle pressure plotting submodule extracts the corresponding time point and quantity value of fund release events based on the cycle number in the high pressure release cycle segment identification table, constructs the release frequency sequence of each cycle unit, arranges the frequency sequence in ascending order by cycle number, matches its position interval in the total cycle, draws a visualization graphic of the release intensity under the corresponding cycle number, and establishes a cycle fund pressure distribution map. Based on the cycle number marked in the high-pressure release cycle segment identification table, for example, cycle number HP-01 [day 14, day 15], extract the corresponding fund release event time points and quantity values. From the analysis of the release rhythm identification submodule, we obtain: 0 releases on day 14, 2 releases on day 15. Construct the release frequency sequence for this cycle unit (day) as [0, 2]. Assume there is another marked high-pressure cycle HP-02 [day 25, day 26], whose corresponding release frequency sequence is [0, 2] (assuming 0 releases on day 25, 2 releases on day 26 [F-03, [F-08] Arrange the frequency sequences of HP-01 and HP-02 in ascending order according to the cycle number (i.e., the chronological order). Match the position interval of HP-01 (days 14 to 15) and the position interval of HP-02 (days 25 to 26) in the total project cycle. Draw a visualization of the release intensity under the corresponding cycle number. Specifically, create a bar chart with time (unit: days) as the horizontal axis and release frequency (unit: number of times) as the vertical axis. Draw a bar with a height of 0 at the horizontal axis "14" (or do not draw one), a bar with a height of 2 at the horizontal axis "15", a bar with a height of 0 at the horizontal axis "25", and a bar with a height of 2 at the horizontal axis "26". Draw the bar charts of all high-pressure cycles on the same chart to establish a cycle capital pressure distribution chart.

[0041] The full-cycle decision generation module includes: The phase combination extraction submodule, based on the period segment number in the periodic capital pressure distribution map, detects whether there is an overlap between the peak of capital release frequency and the task arrangement. It extracts the period number that meets the dual conditions of concentrated release and overlapping tasks, records the associated task number and capital number under each period, establishes the correspondence between period and task capital combination, and generates a concentrated phase combination list. Based on the period segment number HP-01 [Day 14, Day 15] in the periodic capital pressure distribution chart, it is detected whether there is an overlap between the peak frequency of capital release and the task arrangement within this interval. According to the visualization in paragraph 12, there is a frequency peak of 2 on "Day 15". According to the calculation in paragraph 11, the task arrangement density coefficient of this interval is 1.0. The peak threshold is set to 1.5 (i.e., greater than 1.5 times), and the density threshold is set to 0.7. The judgment is executed: (frequency peak 2 > threshold 1.5) and (density coefficient 1.0 > threshold 0.7). Both conditions are met. Therefore, the period number HP-01 that meets the conditions is extracted, and the capital numbers {F-02, F-05} associated with this period (Day 15) and the active task numbers {T-002, T-006, T-007} in this period are recorded. The correspondence between the period and the task capital combination is established, in the format {HP-01: {Tasks: [T-002,T-006, For HP-02 [Day 25, Day 26], perform the same operation, assuming its peak value is 2 (Day 26) and its density is 0.8, which also meets the conditions. Record its associated tasks {T-003, T-008} and funds {F-03, F-08}, and store them in the corresponding relationship {HP-02: ...}. Summarize all the period information that meets the conditions and generate a centralized phase combination list.

[0042] The instruction node integration submodule calls the set of task numbers and fund numbers in the centralized stage combination list, obtains the time tag of the task and the corresponding fund release priority identifier, groups the fund release behavior within the stage according to the start and end time of the task, integrates the path sequence of fund release behavior in chronological order, constructs the combination structure of task segment and fund behavior segment, and obtains the stage task fund path mapping table. Call the HP-01 cycle entry in the centralized phase combination list to obtain the task number set {T-002, T-006, T-007} and the fund number set {F-02, F-05}. Obtain the task time tags (from Table 1 and supplementary data): T-002 [10, 20], T-006 [12, 16], T-007 [14, 17], and obtain the fund release priority identifier (from paragraph 8 and supplementary data): F-02's priority is 2, assuming F-05's priority is 4. Group the fund release behavior within the phase according to the task start and end time. F-02 and F-05 are both released on "day 15". On day 15, tasks T-002, T-006, and T-007 are all within the execution interval. Therefore, F-02 and F-05 are grouped together and released according to the fund release priority identifier (2, 4) Integrate the sequence of fund release behavior paths within the group to obtain F-02 -> F-05, and construct a combined structure of task segment {T-002, T-006, T-007} and fund behavior segment {F-02 -> F-05}. This structure shows that in the HP-01 stage, the concurrent execution of these three tasks corresponds to the concentrated release of F-02 and F-05, thus obtaining the stage task fund path mapping table.

[0043] The decision basis summary submodule identifies the sequence of fund release nodes in each stage based on the task fund segment combination structure in the stage task fund path mapping table, and standardizes it into a standard instruction format. It summarizes the task number, fund number and sequence number fields of all stage combinations and integrates them into a continuous executable data structure sequence to establish a full-cycle financing decision basis set. Based on the HP-01 stage task fund segment combination structure in the stage task fund path mapping table, the node sequence for fund release within this stage is identified as F-02, F-05, and standardized into a standard instruction format. This format includes (stage number, task number, fund number, sequence number). The original association between funds and tasks is queried (Table 2 and supplementary data). F-02 is associated with T-002, with a sequence number of 2. Assuming F-05 is associated with T-006, with a sequence number of 4, the generated standard instructions are: (HP-01, T-002, F-02, 2) and (HP-01, T-006, F-05, 4). The same operation is performed on the HP-02 stage, assuming its sequence is F-03 -> F-08, with associated tasks T-003 and T-008, and sequences 3 and 9 respectively. The generated instructions are: (HP-02, T-003, F-03, 3) and (HP-02, T-008, F-08, 9), summarize all instructions generated from all stages, extract and integrate the task number, fund number and priority number fields from all instructions to form a continuous executable data structure sequence, namely [(HP-01, T-002, F-02, 2), (HP-01, T-006, F-05, 4), (HP-02, T-003, F-03, 3), (HP-02, T-008, F-08, 9)], and establish a full-cycle financing decision basis set.

[0044] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A smart project financing decision support system based on the entire project lifecycle, characterized in that, The system includes: The task link parsing module obtains all task information of the project, identifies the time connection sequence between each task, locates the corresponding events of fund release and task nodes, marks the connection structure between all task nodes and fund behavior, draws the path clue diagram between tasks and funds in sequence, and generates a task and disbursement connection structure diagram. The risk transmission identification module reads the task path in the task and disbursement connection structure diagram, determines whether the sequential relationship between tasks causes a change in the order before and after fund release, identifies the segment where continuous tasks form release interference, records the distribution of interference paths and node sequence positions, and generates a risk transmission distribution map. The fund release sorting module extracts the original time arrangement order based on the high interference nodes identified in the risk transmission distribution map, rearranges the corresponding release time order, updates the node ranking and adjusts the release sorting, and generates a risk priority release order table. The periodic pressure measurement module calls the time period information in the risk priority release sequence table, identifies the changes in task concentration and release distribution during the fund release time period, screens dense release segments, marks time segments with continuous changes in fluctuation amplitude, and generates a periodic fund pressure distribution map.

2. The intelligent project financing decision support system based on the entire life cycle as described in claim 1, characterized in that: The task and disbursement linkage structure diagram includes a task node sequence structure, fund release event nodes, task and fund pairing relationships, and path connection mapping elements. The risk transmission distribution diagram includes a set of delay impact paths, identification of continuous time interference segments, and fund release association chains. The risk priority release order table specifically includes a list of key fund nodes, time order adjustment identifiers, and risk impact priority sequences. The periodic fund pressure distribution diagram includes concentrated release rhythm segments, dense task arrangement intervals, and fund release time pressure points.

3. The intelligent project financing decision support system based on the entire life cycle as described in claim 1, characterized in that, The task link parsing module includes: The task data extraction submodule obtains the task node number, task time tag, and task description field from the task configuration list during the project execution phase. It checks whether the planned start time and end time fields corresponding to each task are complete, calls the task node number and time tag to construct a task time series array, arranges each task number in sequence according to the task time order and marks its corresponding time interval, and generates a task time series structure matrix. The task trigger structure identification submodule determines whether there are overlapping or gap intervals between adjacent task time intervals based on the task number order in the task time sequence structure matrix. If the start time of a task is earlier than the end time of the previous task, it is marked as an overlapping trigger type. If the start time is equal to the end time of the previous task, it is marked as a continuous trigger type. If there is a gap, it is marked as a delayed trigger type. The marking results and task node numbers are used to construct a trigger link logic array and generate a trigger type identification vector between tasks. The task allocation path generation submodule, based on the task number order in the inter-task trigger type identifier vector, calls the fund behavior number associated with the corresponding task node in the fund allocation record table, matches whether the fund allocation time and the task execution time overlap. If the time overlaps, a connection structure is established between the fund behavior number and the task number, constructs a one-to-one mapping matrix between task nodes and fund behavior nodes, draws a path clue diagram composed of fund behavior nodes in the order of task nodes, and generates a task allocation connection structure diagram.

4. The intelligent project financing decision support system based on the entire life cycle as described in claim 3, characterized in that: The process of generating the task time series structure matrix is ​​as follows: creating the task time series structure matrix, which includes a task node number column, a planned start time column, and an end time column; filling the task node number, the planned start time, and the end time into the corresponding columns of the task time series structure matrix, and sorting the task time series structure matrix in ascending order according to the planned start time; The process of matching whether the fund disbursement time and the task execution time coincide is as follows: obtain the fund disbursement time corresponding to the fund behavior number in the fund disbursement record table; extract the planned start time and the end time corresponding to the task number from the task time sequence structure matrix; determine whether the fund disbursement time is within the closed time interval formed by the planned start time and the end time; if the fund disbursement time is within the closed time interval, it is determined that the fund disbursement time and the task execution time coincide.

5. The intelligent project financing decision support system based on the entire life cycle as described in claim 1, characterized in that, The risk transmission identification module includes: The path structure parsing submodule, based on the task path structure in the task and disbursement connection structure graph, obtains the task node number and its corresponding fund release time in each path, detects whether the time interval between adjacent task nodes meets the continuation triggering condition, determines whether the time offset of the preceding node causes the change in the start time of the subsequent node, filters out task combinations with preceding and following relationships, and generates a list of delay impact transmission paths. The interference segment determination submodule calculates the cumulative time offset of each node on the task path based on the time difference of the task pairs in the delay effect transmission path list, compares it with the set continuous interference threshold of task time, determines whether there is a continuous node segment that meets the interference interval length requirement, calls the cumulative time offset sequence and the task node order, marks the formed continuous interference node set, and obtains the continuous interference segment interval vector. The transmission trajectory construction submodule calls the task number combination in the interval vector of the continuous interference segment, matches its associated fund release number and release time point, connects the fund release nodes with continuous task interference relationship, arranges the fund release behavior according to the time sequence, draws the impact extension trajectory on each path, and establishes a risk transmission distribution map.

6. The intelligent project financing decision support system based on the entire life cycle as described in claim 1, characterized in that, The fund release sorting module includes: The key node extraction submodule, based on the path structure in the risk transmission distribution map, detects whether the task node corresponding to the fund release node in the path is within the marked range of the continuous interference segment interval vector, extracts the set of task numbers that meet the continuous interference condition, obtains their original fund release time label, and forms a mapping relationship with the task number to generate an interference key node time index table. The release priority adjustment submodule calls the original time tag sequence of the fund release nodes in the time index table of the interference key nodes, sorts them in ascending order according to the time sequence, and reassigns a new priority identifier to each fund release node. At the same time, it constructs a difference matrix between the old priority and the new priority, establishes a release priority adjustment structure, and obtains a fund release priority rearrangement matrix. The sorting structure generation submodule recombines the fund release number with its adjusted order index according to the new order identifier in the fund release order rearrangement matrix, generates a complete list of release numbers and a corresponding list of release times, constructs an updated time sorting mapping table corresponding to the risk path structure, and establishes a risk priority release order table.

7. The intelligent project financing decision support system based on the entire life cycle as described in claim 1, characterized in that, The periodic pressure measurement module includes: The release rhythm identification submodule, based on the fund release time field in the risk priority release sequence table, counts the number of fund release events in each time period, calculates the release frequency difference between adjacent time periods, determines whether the release frequency change exceeds the set release frequency fluctuation threshold, filters out time segments that continuously exceed the fluctuation threshold, and generates a list of concentrated fund release intervals. The dense section determination submodule calls the time period information in the list of concentrated fund release intervals to obtain the task node number and its arrangement time interval within the corresponding period. It calculates the task arrangement density coefficient based on the overlap of the time intervals, compares the arrangement density coefficient with the number of fund release events in parallel, and filters out the period positions where tasks are dense and release behavior occurs continuously to obtain the high-pressure release period segment identification table. The cycle pressure plotting submodule extracts the corresponding time point and quantity value of fund release events based on the cycle number in the high-pressure release cycle segment identifier table, constructs the release frequency sequence for each cycle unit, arranges the frequency sequence in ascending order by cycle number, matches its position interval in the total cycle, plots a visualization graphic of the release intensity under the corresponding cycle number, and establishes a cycle fund pressure distribution map.

8. The intelligent project financing decision support system based on the entire life cycle as described in claim 7, characterized in that: The process of counting the number of fund release events in each time period is as follows: set the statistical period unit to natural day; iterate through the fund release time field in the risk priority release order table and count the number of fund release events falling within the same natural day. The process of calculating the task arrangement density coefficient based on the overlap of time intervals is as follows: obtain all the task node numbers corresponding to the time period in the list of concentrated fund release intervals; extract the arrangement time intervals corresponding to all the task node numbers, calculate the total duration of all the arrangement time intervals within the time period, and calculate the overlap duration of any two arrangement time intervals within the time period; sum the overlap durations, and divide the sum of the overlap durations by the total duration to obtain the task arrangement density coefficient.

9. The intelligent project financing decision support system based on the entire life cycle as described in claim 1, characterized in that, The system also includes: The full-cycle decision generation module reads the pressure release segment in the cycle capital pressure distribution map, extracts the relationship between task distribution and release rhythm, identifies the task order rearrangement and combination, organizes the corresponding structure of task segments and capital release time lines, and generates a full-cycle financing decision basis set. The full-cycle financing decision-making basis set includes stage combination classification items, key task time sequence position, and task funding path overview table.

10. The intelligent project financing decision support system based on the entire life cycle as described in claim 9, characterized in that, The full-cycle decision generation module includes: The phase combination extraction submodule, based on the period segment number in the periodic capital pressure distribution map, detects whether there is an overlapping interval between the peak of capital release frequency and the task arrangement, extracts the period number that meets the dual conditions of concentrated release and overlapping tasks, records the associated task number and capital number under each period, establishes the correspondence between period and task capital combination, and generates a concentrated phase combination list. The instruction node integration submodule calls the set of task numbers and fund numbers in the centralized stage combination list, obtains the time tag of the task and the corresponding fund release priority identifier, groups the fund release behavior within the stage according to the start and end time of the task, integrates the path sequence of fund release behavior in chronological order, constructs the combination structure of task segment and fund behavior segment, and obtains the stage task fund path mapping table. The decision basis summary submodule identifies the sequence of fund release nodes in each stage based on the task fund segment combination structure in the stage task fund path mapping table, standardizes it into a standard instruction format, summarizes the task number, fund number and sequence number fields of all stage combinations, integrates them into a continuous executable data structure sequence, and establishes a full-cycle financing decision basis set.

Images