Decision analysis method and system for standard material purchasing system

Abstract

The invention discloses a decision analysis method and system for a standard material purchasing system, and relates to the technical field of purchasing decision, and the method comprises the steps: collecting a historical material purchasing data flow, extracting a sum value, a supplier credit rating, and logistics cost, and combining the sum value, the supplier credit rating, and the logistics cost into a basic feature set; carrying out big data analysis attenuation modeling on the aging attribute of the basic feature set, establishing a hyperbolic attenuation model and outputting a dynamic attenuation factor; mapping the basic feature set into a coordinate vector, inputting the coordinate vector and a dynamic attenuation factor into a Lorentz compression engine, and compressing and storing the coordinate vector and the dynamic attenuation factor into a quotation memory bank; and converting the purchasing scheme into an electronic order, distributing the electronic order to a supplier platform, collecting performance time delay data, and calibrating a dynamic attenuation factor to complete a closed loop. According to the method, dynamic modeling is carried out on the aging attribute of the basic feature set through the hyperbolic attenuation model, and the dynamic attenuation factor is enabled to be adaptive to the market environment change by using a nonlinear coupling mechanism of the market fluctuation rate and the logistics cost drift factor.

Description

Technical Field

[0001] This invention relates to the field of procurement decision-making technology, and in particular to a decision analysis method and system for a standard material procurement system. Background Technology

[0002] In recent years, the supply chain management field has widely adopted big data analytics to optimize procurement decisions, with significant progress made, particularly in supplier evaluation and price analysis. Machine learning-based cost prediction models, blockchain-enabled credit rating, and real-time logistics tracking technologies have enabled the structured processing of multi-source procurement data. Meanwhile, the application of quantum computing in data compression provides a new path for storing and retrieving massive amounts of historical price data, while thermodynamic optimization algorithms have demonstrated theoretical advantages in resolving multi-objective decision-making conflicts.

[0003] Existing technologies for modeling the decay of the time-sensitive value of historical price data have shortcomings. Traditional exponential decay models use fixed decay coefficients, which cannot dynamically adapt to the non-linear changes in time-sensitive value caused by market fluctuations. This static modeling approach leads to a large amount of data in the historical price database becoming invalid due to time-sensitive distortion, and this cannot be resolved by adjusting conventional parameters. It also ignores the coupling effect of supply chain risk transmission, logistics time-sensitive drift, and the dynamics of credit rating, resulting in a serious deviation between the data foundation on which procurement decisions are based and the actual value. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a decision analysis method for a standard material procurement system to solve the problem of dynamic quantification and cross-cycle calibration of the time-sensitive value of historical quotation data in procurement decisions.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0007] In a first aspect, the present invention provides a decision analysis method for a standard material procurement system, which includes collecting historical material procurement data streams and extracting monetary values, supplier credit ratings, and logistics cost combinations as a basic feature set.

[0008] Big data analysis and attenuation modeling are performed on the time-related attributes of the basic feature set to establish a hyperbolic attenuation model and output a dynamic attenuation factor;

[0009] The basic feature set is mapped to a coordinate vector, and the dynamic decay factor is input into the Lorentz compression engine for compressed storage as a quotation memory.

[0010] Receive procurement instructions, identify the material code parameters in the procurement instructions, retrieve supplier records that meet the time and space constraints in the quotation memory, and generate a candidate quotation set by sorting the quotation value.

[0011] Receive cost parameters, budget parameters, and supply chain risk parameters submitted based on the candidate quotation set, and convert them into a three-dimensional decision vector by reducing their dimensionality. Use the thermodynamic entropy reduction algorithm to converge the angle of the three-dimensional decision vector and output the procurement plan with an entropy value lower than the preset entropy threshold.

[0012] The procurement plan is converted into electronic orders and distributed to the supplier platform. Performance delay data is collected, and the dynamic attenuation factor is calibrated to complete the closed loop.

[0013] As a preferred embodiment of the decision analysis method for the standard material procurement system described in this invention, the combination is a basic feature set, and the specific steps are as follows:

[0014] The system uses a regular expression parsing engine to scan the structured fields of historical material procurement data streams, extracting the monetary value field, supplier credit rating field, and logistics cost field to generate an original feature set.

[0015] Apply hyperbolic time-decay weighting to the timestamp attributes of the original feature set to generate a time-weighted feature set.

[0016] The time-weighted feature set is input into the Pareto encoder for three-dimensional vector normalization mapping to generate a standard feature vector set, which is then stored as the basic feature set.

[0017] As a preferred embodiment of the decision analysis method for the standard material procurement system described in this invention, the specific steps for establishing a hyperbolic attenuation model and outputting a dynamic attenuation factor are as follows:

[0018] Using the monetary value in the basic feature set as the search key, perform a matching query on the amount with a tolerance lower than the preset tolerance threshold in the historical purchase order database to obtain the order identifier;

[0019] The actual transaction price is obtained by querying historical purchase fulfillment records using order identifiers. Based on the ratio of the actual transaction price to the supplier's initial quotation, an actual price validity sequence is generated.

[0020] Market volatility is calculated using supplier credit ratings based on a set of basic features, and benchmark validity is estimated using logistics costs. A hyperbolic decay model is established and optimized by inputting the actual price validity sequence into a quantum annealer, outputting a dynamic decay factor.

[0021] As a preferred embodiment of the decision analysis method for the standard material procurement system described in this invention, the compressed storage is a quotation memory, and the specific steps are as follows:

[0022] The monetary values, supplier credit ratings, and logistics costs of the basic feature set are normalized and mapped into coordinate vectors, and a four-dimensional spatiotemporal coordinate vector is constructed by combining the dynamic decay factor.

[0023] Based on market volatility, the Lorentz transformation velocity parameters are calculated, and the four-dimensional spatiotemporal coordinate vector is spatiotemporally compressed to generate a compressed state.

[0024] The compressed state is input into the quantum error correction encoder, and logical qubits are generated using the surface code algorithm.

[0025] Logical qubits are written into quantum memory cells and associated with suppliers to form a quotation memory bank.

[0026] As a preferred embodiment of the decision analysis method for the standard material procurement system described in this invention, the specific steps for generating a candidate price set by sorting prices by price value are as follows:

[0027] Receive procurement instructions, parse the material codes, procurement quantities, and urgency parameters in the procurement instructions, and generate demand quantum states through quantum amplitude encoding;

[0028] Extract supplier quantum states from the quotation memory and construct a dynamic hypergraph network based on logistics synergy and competition intensity;

[0029] The demand quantum state and dynamic hypergraph network are input into the quantum field evolution equation, and the supplier matching degree is calculated through the interaction Hamiltonian.

[0030] Based on the matching degree between the quoted price and the supplier, a hyperbolic geometric sorting is performed in the Poincaré disk model to generate a set of candidate quotes.

[0031] As a preferred embodiment of the decision analysis method for the standard material procurement system described in this invention, the specific steps for determining procurement schemes with output entropy values ​​lower than a preset entropy threshold are as follows:

[0032] The cost parameters, budget parameters, and supply chain risk parameters submitted based on the candidate bid set will be mapped into complex quantum fields to generate quantum resonant ground states;

[0033] The quantum resonant ground state is input into the Chern-Simons topological field to construct the decision connection field, and the quantum vortex dynamics equation is solved to output the vortex field.

[0034] The vortex field is evolved through quantum resonance tunneling conditions, and the spatial expectation value is calculated to output a three-dimensional decision vector;

[0035] Calculate the entropy value of the three-dimensional decision vector. If the entropy value is lower than the preset entropy threshold, output the procurement plan. If the entropy value is higher than the preset entropy threshold, perform quantum annealing optimization and regenerate the quantum resonant ground state.

[0036] As a preferred embodiment of the decision analysis method for the standard material procurement system described in this invention, the calibration of the dynamic attenuation factor completes a closed loop, and the specific steps are as follows.

[0037] The procurement plan is encoded into quantum state order packages and distributed to the supplier platform, and performance delay data is collected;

[0038] Construct a time delay curvature model based on the collected performance delay data, and output a curvature scalar;

[0039] The curvature scalar is input into the quantum phase-locked engine to generate phase gradient data, which is then used to adjust the historical dynamic decay factor and generate a calibrated dynamic decay factor.

[0040] The Lorentz compression engine, which feeds back the calibrated dynamic decay factor to the quotation memory, completes the real-time update.

[0041] Secondly, the present invention provides a decision analysis system for a standard material procurement system, including a feature acquisition module, an attenuation modeling module, a quantum compression module, a supplier screening module, an entropy reduction decision module, and a closed-loop calibration module;

[0042] The feature acquisition module is used to collect historical material procurement data streams and extract the basic feature set based on the amount value, supplier credit rating, and logistics cost combination.

[0043] The attenuation modeling module is used to perform big data analysis and attenuation modeling on the time-related attributes of the basic feature set, establish a hyperbolic attenuation model, and output a dynamic attenuation factor.

[0044] The quantum compression module is used to map the basic feature set into a coordinate vector, and input the dynamic decay factor into the Lorentz compression engine for compression and storage as a quotation memory.

[0045] The supplier screening module is used to receive procurement instructions, identify material code parameters in procurement instructions, retrieve supplier records that meet the time and space constraints in the quotation memory, and generate a candidate quotation set by sorting the quotation value.

[0046] The entropy reduction decision module is used to receive cost parameters, budget parameters and supply chain risk parameters submitted based on the candidate quotation set, and convert them into a three-dimensional decision vector. The thermodynamic entropy reduction algorithm is used to converge the angle of the three-dimensional decision vector and output a procurement plan with an entropy value lower than a preset entropy value threshold.

[0047] The closed-loop calibration module is used to convert the procurement plan into electronic orders and distribute them to the supplier platform, collect performance delay data, and calibrate the dynamic attenuation factor to complete the closed loop.

[0048] Thirdly, the present invention provides a computer device including a memory and a processor, wherein the memory stores a computer program, wherein when the computer program is executed by the processor, it implements any step of the decision analysis method of the standard material procurement system as described in the first aspect of the present invention.

[0049] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, it implements any step of the decision analysis method of the standard material procurement system as described in the first aspect of the present invention.

[0050] The beneficial effects of this invention are as follows: It dynamically models the timeliness attributes of the basic feature set using a hyperbolic decay model, and leverages the nonlinear coupling mechanism between market volatility and logistics cost drift factors to enable the dynamic decay factor to autonomously adapt to changes in the market environment. The hyperbolic decay model uses supplier credit ratings as a dynamic adjustment variable, combined with industry benchmark logistics costs, to achieve precise control of the decay coefficient, thereby improving the fidelity of the timeliness value of historical price data. Attached Figure Description

[0051] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0052] Figure 1 This is a flowchart of the decision analysis method for a standard materials procurement system.

[0053] Figure 2 This is a module diagram of the decision analysis system for the standard materials procurement system.

[0054] Figure 3 This is a flowchart for combining the basic feature sets.

[0055] Figure 4 This is a flowchart for outputting the procurement plan. Detailed Implementation

[0056] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0057] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0058] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0059] Reference Figures 1-4 As an embodiment of the present invention, this embodiment provides a decision analysis method for a standard material procurement system, including the following steps:

[0060] S1. Collect historical material procurement data streams and extract the basic feature set based on the amount, supplier credit rating, and logistics cost combination.

[0061] Furthermore, the structured fields of historical material procurement data streams are scanned using a regular expression parsing engine to extract numerical value fields, supplier credit rating fields, and logistics cost fields to generate an original feature set;

[0062] Specifically, the regular expression parsing engine scans the structured text content of historical material procurement data streams, identifies regular expression patterns in the monetary value field to match currency symbols and number combinations, captures letter rating patterns in the supplier credit rating field, detects associated values ​​in the logistics cost field, and aligns the extracted monetary value field, supplier credit rating field, and logistics cost field according to the record timestamp to generate the original feature set.

[0063] Apply hyperbolic time-decay weighting to the timestamp attributes of the original feature set to generate a time-weighted feature set.

[0064] Specifically, the timestamp attribute is extracted from the original feature set and converted into a standard time format. The time interval difference is obtained based on the timestamp in the standard time format and the current time. The time interval difference is input into the hyperbolic time-delay function, which outputs the time-delay weight coefficient. The time-delay weight coefficient is combined with the monetary value field of the original feature set to generate a time-delay weighted monetary value. The time-delay weight coefficient is combined with the supplier credit rating field of the original feature set to generate a time-delay weighted supplier credit rating. The time-delay weight coefficient is combined with the logistics cost field of the original feature set to generate a time-delay weighted logistics cost. The time-delay weighted monetary value, the time-delay weighted supplier credit rating, and the time-delay weighted logistics cost are combined to generate a time-delay weighted feature set.

[0065] The time-weighted feature set is input into the Pareto encoder for three-dimensional vector normalization mapping to generate a standard feature vector set, which is then stored as the basic feature set.

[0066] Specifically, the time-weighted feature set is input into the Pareto encoder to perform three-dimensional vector normalization mapping. The Pareto encoder maps the ratio of the time-weighted amount to the historical highest price of similar materials to the X-axis coordinate. The credit scaling factor is obtained by querying the numerical coefficient corresponding to the supplier credit rating field in the preset mapping table. The time-weighted supplier credit rating and the credit scaling factor are combined and mapped to the Y-axis coordinate. The time-weighted logistics cost is mapped to the Z-axis coordinate by taking the natural logarithm. The generated X-axis coordinate, Y-axis coordinate, and Z-axis coordinate are combined to form a standard feature vector set. The standard feature vector set is written into the basic feature set.

[0067] It should be noted that the preset mapping table is set based on the median decision weight of suppliers with different credit ratings in the industry's historical procurement data, such as 0.8 for grade A, 1.0 for grade B, and 1.2 for grade C.

[0068] S2. Perform big data analysis and attenuation modeling on the time-related attributes of the basic feature set, establish a hyperbolic attenuation model, and output the dynamic attenuation factor.

[0069] Furthermore, using the monetary value in the basic feature set as the search key, a matching query is performed in the historical purchase order database for amounts with a tolerance lower than the preset tolerance threshold to obtain the order identifier;

[0070] It should be noted that the preset tolerance threshold is set based on the statistical fluctuation range of the amount field in the historical purchase order database, with the example value being one percent.

[0071] Specifically, the amount value in the basic feature set is used as the search key to input the historical purchase order database. The historical purchase order database defines an amount matching rule with a tolerance lower than the preset tolerance threshold. The amount matching query is executed to scan the order amount field, filter order records that meet the tolerance lower than the preset tolerance threshold, and extract the order identifier of the matching order record. When the amount tolerance is higher than the preset tolerance threshold, the order identifier query terminates and returns an empty result set.

[0072] The actual transaction price is obtained by querying historical purchase fulfillment records using order identifiers. Based on the ratio of the actual transaction price to the supplier's initial quotation, an actual price validity sequence is generated.

[0073] Specifically, a structured query is performed on historical purchase fulfillment records using order identifiers. The historical purchase fulfillment records return the actual transaction price associated with the order identifier. The initial quote from the supplier corresponding to the same order identifier is extracted, and the ratio of the actual transaction price to the initial quote from the supplier is calculated. This ratio serves as the price validity value, and the price validity values ​​are arranged in order of order timestamp to form an actual price validity sequence.

[0074] Market volatility is calculated using supplier credit ratings based on a set of basic features, and benchmark validity is estimated using logistics costs. A hyperbolic decay model is established and optimized by inputting the actual price validity sequence into a quantum annealer, outputting a dynamic decay factor.

[0075] Specifically, the hyperbolic decay model and the actual price-effectiveness sequence are input into the quantum annealer. The quantum annealer encodes the parameters of the hyperbolic decay model into qubit spin states, and the actual price-effectiveness sequence is transformed into the energy function of the Ising model. The quantum annealer performs an annealing evolution process from high temperature to low temperature. The evolution process solves for the global minimum of the Ising model energy function. The qubit state corresponding to the global minimum is decoded into optimization parameters. The optimization parameters are substituted into the hyperbolic decay model to output the dynamic decay factor.

[0076] Market volatility is calculated using supplier credit ratings from a set of basic features, and the benchmark validity period is extrapolated using logistics costs. A hyperbolic decay model is then established, with the following expression:

[0077]

[0078] In the formula, V(t) represents the price validity value at the current time t, β represents the dynamic decay coefficient, t represents the current time, σ represents market volatility, k represents the credit scaling factor, τ represents the benchmark validity period, and σ min σ represents the minimum market volatility. max τ represents the highest market volatility, k1 represents the credit sensitivity coefficient, C represents the supplier credit rating, τ0 represents the benchmark validity period, L represents logistics costs, L0 represents the minimum logistics costs, and α represents the logistics cost elasticity index.

[0079] It should be noted that the dynamic decay coefficient is calculated from market volatility, credit scaling factor, and benchmark validity period, with an example value of 0.0073; the credit scaling factor is based on a linear mapping of supplier credit rating, with an example value of 0.8 for A-level suppliers (rewarding low risk); the credit sensitivity coefficient is set to a fixed value based on the industry's credit risk transmission strength, with an example value of 0.8 for the electronic component procurement industry benchmark; and the logistics cost elasticity index is fitted to the marginal effect of logistics costs on validity period using historical data, with an example value of 0.7.

[0080] S3. Map the basic feature set to coordinate vectors, input them along with the dynamic decay factor into the Lorentz compression engine, and compress and store them as a quotation memory.

[0081] Furthermore, the monetary values, supplier credit ratings, and logistics costs of the basic feature set are normalized and mapped into coordinate vectors, and a four-dimensional spatiotemporal coordinate vector is constructed by combining the dynamic decay factor.

[0082] Specifically, the ratio of the monetary value of the basic feature set to the historical highest price of similar materials is normalized to the X-axis component, the ratio of the supplier credit rating to the credit scaling factor is normalized to the Y-axis component, and the logistics cost is normalized to the natural logarithm to the Z-axis component. The normalized X-axis component, normalized Y-axis component, and normalized Z-axis component are combined to form a coordinate vector, and the coordinate vector is combined with the dynamic decay factor to form a four-dimensional spatiotemporal coordinate vector.

[0083] Based on market volatility, the Lorentz transformation velocity parameters are calculated, and the four-dimensional spatiotemporal coordinate vector is spatiotemporally compressed to generate a compressed state.

[0084] Specifically, the Lorentz transformation velocity parameter is calculated based on market volatility, generating a Lorentz transformation matrix. The Lorentz transformation matrix is ​​applied to the four-dimensional spacetime coordinate vector to perform spatial contraction and time dilation transformations. The Lorentz contraction factor is dynamically generated through the relativistic velocity parameter converted from market volatility. The spatial contraction transformation combines the spatial components of the four-dimensional spacetime coordinate vector with the Lorentz contraction factor to generate the transformed spatial components. The time dilation transformation combines the time components of the four-dimensional spacetime coordinate vector with the Lorentz dilation factor to generate the transformed time components. The transformed spatial components and the transformed time components are combined to form a compressed state.

[0085] The Lorentz transform velocity parameter is calculated based on market volatility, and the expression is as follows:

[0086]

[0087] In the formula, ε represents the Lorentz transform velocity parameter, ε max σ represents the maximum relativistic velocity, tanh represents the hyperbolic tangent function, and σ represents the maximum relativistic velocity. crit This represents the root mean square error of historical market volatility.

[0088] The compressed state is input into the quantum error correction encoder, and logical qubits are generated using the surface code algorithm.

[0089] Specifically, the compressed state input quantum error correction encoder is loaded into the initial quantum state. The quantum error correction encoder calls the surface code algorithm to define the qubit lattice structure. The surface code algorithm initializes the physical qubit array on the lattice. The physical qubit array performs a stabilizer measurement operation. The stabilizer measurement result performs error detection. The detected error position triggers the error correction gate operation. The error-corrected physical qubit state is initialized into a logical state. The logical state initialization process generates logical qubits and maintains quantum coherence. The surface code algorithm verifies the fidelity of the logical qubits. When the preset fidelity threshold is reached, the logical qubits are output.

[0090] It should be noted that the preset fidelity threshold is based on the theoretical fault tolerance setting of quantum error correction codes, and the example value is 0.98; if the preset fidelity threshold is not reached, the quantum state will be reinitialized and the error correction will be repeated.

[0091] Logical qubits are written into quantum memory cells and associated with suppliers to form a quotation memory bank.

[0092] Specifically, logical qubits are input into quantum memory units to perform quantum state writing operations. The quantum memory units are allocated physical storage addresses, and a mapping is established between the physical storage addresses and the supplier's unique identifier. The supplier's unique identifier queries the supplier's master database to obtain supplier metadata, which includes the supplier's name, the type of supplied materials, and historical performance records. The supplier's metadata is bound to the physical address of the logical qubits, and the bound supplier metadata is encapsulated into a quotation memory record. The quotation memory record is stored in order of timestamps and supports quantum state deserialization call interfaces. The supplier association process is fast-retrieval through hash indexes, forming the quotation memory.

[0093] S4. Receive the procurement instruction, identify the material code parameters in the procurement instruction, retrieve supplier records that meet the time and space constraints in the quotation memory, and generate a candidate quotation set by sorting the quotation value.

[0094] Furthermore, it receives procurement instructions, parses the material codes, procurement quantities, and urgency parameters in the procurement instructions, and generates demand quantum states through quantum amplitude encoding;

[0095] Specifically, the material code is mapped to the quantum ground state selection basis, the procurement quantity is normalized to the ground state amplitude ratio, the urgency is converted into the phase angle parameter, and the ground state amplitude ratio and phase angle parameter are input into the quantum amplitude encoder to generate the demand quantum state.

[0096] Extract supplier quantum states from the quotation memory and construct a dynamic hypergraph network based on logistics synergy and competition intensity;

[0097] Specifically, the supplier quantum states stored in the price quote memory are extracted and deserialized. The deserialized supplier quantum states serve as hypergraph nodes. The proportion of shared logistics routes among suppliers is obtained based on logistics synergy, and the price difference index among suppliers is obtained based on competition intensity. The combination of logistics synergy and competition intensity defines the hypergraph hyperedge weights. The hyperedge weights connect and associate supplier nodes. The dynamic hypergraph network updates the hyperedge weights based on the real-time shared logistics route proportion and price difference index, thus generating the dynamic hypergraph network.

[0098] Specifically, the demand quantum state and the dynamic hypergraph network are input into the quantum field evolution equation, and the supplier matching degree is calculated through the interaction Hamiltonian, expressed as:

[0099]

[0100] In the formula, Let represent the matching degree of supplier i, where i represents the supplier index, and T represents the procurement decision cycle. Indicates the reciprocal of the procurement decision-making cycle. Indicates the integration from 0 to T, e -Γt Let Γ represent the exponential decay term, Γ represent the environmental decoherence rate, and <φ| represent the left vector of the demand quantum state. Let r denote the time evolution operator, r denote the imaginary unit, and H denote the interaction Hamiltonian. Denotes the reduced Planck constant. Denotes the right vector of the hypergraph quantum state of supplier i, |·| 2 dt represents the squared modulus of quantum state fidelity, and dt represents the time differential element.

[0101] Based on the matching degree between the quoted price and the supplier, a hyperbolic geometric sorting is performed in the Poincaré disk model to generate a set of candidate quotes.

[0102] It should be noted that the pre-training process of the Poincaré disk model is as follows: the historical procurement dataset is input into the hyperbolic embedding layer to initialize the coordinate parameters, historical supplier relationship network data is collected, a basic graph structure is constructed, the geodesic distance between nodes is extracted from the basic graph structure as a supervision signal, the hyperbolic embedding layer performs Riemann gradient descent optimization, the loss function minimizes the difference between the predicted distance and the actual geodesic distance, the optimization process iteratively updates the node embedding coordinates, the node embedding coordinate update satisfies the hyperbolic space parallel translation constraint, after convergence, the hyperbolic embedding layer parameters are fixed, the trained Poincaré disk model stores the curvature constant and the embedding matrix, and finally the Poincaré disk model is obtained.

[0103] Specifically, the bid value is mapped to the radial coordinates of the Poincaré disk model, and the supplier matching degree is mapped to the angular coordinates. The bid value and supplier matching degree are combined to define the coordinate points of the Poincaré disk. The hyperbolic distance between each coordinate point is calculated. The hyperbolic distance reflects the overall priority of the bid. The supplier records are sorted in ascending order of hyperbolic distance, and the sorting results generate a candidate bid set.

[0104] S5. Receive cost parameters, budget parameters, and supply chain risk parameters submitted based on the candidate quotation set, and convert them into a three-dimensional decision vector by reducing their dimensionality. Use a thermodynamic entropy reduction algorithm to converge the angle of the three-dimensional decision vector and output a procurement plan with an entropy value lower than the preset entropy threshold.

[0105] Furthermore, the cost parameters, budget parameters, and supply chain risk parameters submitted by the candidate bid set will be mapped into complex quantum fields to generate quantum resonant ground states;

[0106] Specifically, the cost parameters submitted based on the candidate bid set are mapped to the real components of the complex quantum field, the budget parameters are mapped to the imaginary components, and the supply chain risk parameters are converted into phase angles. The combination of the real components, imaginary components, and phase angles defines the complex quantum field. The complex quantum field is input into the quantum harmonic oscillator generator to solve the ground state solution of the Schrödinger equation. The ground state solution is normalized and used as the quantum harmonic ground state.

[0107] Specifically, the quantum resonant ground state is input into the Chern-Simons topological field to construct the decision connection field, and the quantum vortex dynamics equation is solved to output the vortex field, which is expressed as:

[0108]

[0109] In the formula, Indicates the location The vortex field vector at that location, Represents a position vector. Represents the curl operator, The vector represents the decision-making connection field, and γ represents the nonlinear coupling coefficient. This represents a scalar field representing the self-interaction of vortices. Let Φ represent the expectation value of the quantum resonant ground state under the topological Hamiltonian gradient, and let Φ represent the quantum resonant ground state. Represents the gradient operator. Represents the topological Hamiltonian. Let Φ denote the topological Hamiltonian gradient, <Φ| denote the left vector of the quantum resonant basis, and |Φ> denote the right vector of the quantum resonant basis.

[0110] Specifically, the vortex field is evolved through quantum resonance tunneling conditions, and the spatial expectation value is calculated to output a three-dimensional decision vector, expressed as:

[0111]

[0112] In the formula, Represents a three-dimensional decision vector. This represents the integral operation of the vortex field vector in the three-dimensional decision space, where Δt represents the decision time window, and e -ΓΔt The term represents the exponential decay term, and <Ψ| represents the left vector of the tunneling reference state. Denotes the resonant tunneling Hamiltonian, and |Φ> denotes the right vector of the vortex field quantum state. This indicates that the quantum state of the vortex field evolves to the tunneling reference state under the action of the resonant tunneling Hamiltonian. Represents a three-dimensional volume element, ∫e -ΓΔt This represents the normalized denominator.

[0113] Calculate the entropy value of the three-dimensional decision vector. If the entropy value is lower than the preset entropy threshold, output the procurement plan. If the entropy value is higher than the preset entropy threshold, perform quantum annealing optimization and regenerate the quantum resonant ground state.

[0114] It should be noted that the preset entropy threshold is set based on the median of the Shannon entropy distribution of historical purchasing decisions, with an example value of 0.3.

[0115] Specifically, the entropy value of the three-dimensional decision vector is compared with a preset entropy threshold. If the entropy value is lower than the preset entropy threshold, the procurement plan is directly output. If the entropy value is higher than the preset entropy threshold, the quantum annealing optimization process is started. The quantum annealing optimization process reconstructs the Hamiltonian parameters. The Hamiltonian parameters are input into the quantum annealer to solve for the ground state. The ground state is decoded into optimization decision parameters. The optimization decision parameters regenerate the quantum resonant ground state. The new quantum resonant ground state returns to the starting point of the decision process.

[0116] The entropy value of the three-dimensional decision vector is calculated using the following expression:

[0117]

[0118]

[0119] In the formula, The entropy value represents the three-dimensional decision vector. Let ln denote the normalization factor, and ln denote the natural logarithm function. This represents summing over all decision components, where j represents the index of the decision component, and λ represents the summation over all decision components. j Let lnλ represent the activation probability of the j-th decision component. j e represents the logarithm of the activation probability of the j-th decision component. -δ(Dj- μ j ) represents the exponential decision bias term, and δ represents the decision sharpness coefficient. μ represents the j-th decision component of the three-dimensional decision vector. j This represents the baseline threshold for the j-th decision component.

[0120] It should be noted that the normalization factor is derived from the maximum entropy constraint of the three states in information theory, and the example value is fixed at 0.910; the decision sharpness coefficient is derived from the supply chain stability quantification, and the example value is 8.1; the benchmark thresholds of the decision components are set based on the median of historical data, and the example values ​​are: cost component μ1 = 0.6 (cost-sensitive threshold), budget component μ2 = 1.0 (budget saturation threshold), and risk component μ3 = 0.4 (risk warning threshold).

[0121] S6. Convert the procurement plan into electronic orders and distribute them to the supplier platform, collect performance delay data, and calibrate the dynamic attenuation factor to complete the closed loop.

[0122] Furthermore, the procurement plan is encoded into quantum state order packages and distributed to the supplier platform, and performance delay data is collected;

[0123] Specifically, the procurement plan inputs the quantum state order encoder, which outputs the cost ratio, procurement quantity, and urgency parameters. The cost ratio is converted into the Y-axis rotating door angle, the procurement quantity is calculated as the ground state amplitude ratio, and the urgency is mapped to the phase angle. The rotating door operation, amplitude ratio, and phase angle are combined to form a quantum state order package. The quantum state order package is distributed to the supplier platform interface through the quantum channel. The supplier platform receives the quantum state order package and returns a quantum state confirmation signal. The quantum state confirmation signal triggers the start of the fulfillment timer. The supplier's shipment operation generates an actual shipment timestamp. The platform's promised delivery timestamp and shipment timestamp are input for delay calculation. The delay data includes the purchase order number, promised time, and shipment time fields. The delay data is written to the historical fulfillment database. At the end of the fulfillment monitoring cycle, the fulfillment delay data is output.

[0124] Construct a time delay curvature model based on the collected performance delay data, and output a curvature scalar;

[0125] Specifically, the promised delivery time and the actual shipment timestamp are extracted from the performance delay data. Based on the timestamp difference between the promised delivery time and the actual shipment timestamp, a time difference vector is generated. The time difference vector is mapped to four-dimensional spacetime coordinates. The four-dimensional spacetime coordinates are input into the Riemann geometry engine, which outputs a Riemann curvature tensor. The Riemann curvature tensor undergoes a scalar curvature contraction operation. After the contraction is completed, a curvature scalar is output.

[0126] The curvature scalar is input into the quantum phase-locked engine to generate phase gradient data, which is then used to adjust the historical dynamic decay factor and generate a calibrated dynamic decay factor.

[0127] Specifically, the curvature scalar is input into the quantum phase-locked engine to drive the frequency modulator. The frequency modulator generates a reference phase field. The interference pattern between the reference phase field and the quantum resonant cavity is input into the phase gradient detection circuit. The phase gradient detection circuit outputs differential phase difference data. The differential phase difference data is used as phase gradient data. The phase gradient data is input into the dynamic decay factor updater. The updater loads the historical dynamic decay factor and performs an exponential weighted update. After the weighted update, the calibrated dynamic decay factor is output.

[0128] The Lorentz compression engine, which feeds back the calibrated dynamic decay factor to the quotation memory, completes the real-time update.

[0129] Specifically, after calibration, the dynamic attenuation factor is transmitted to the input port of the quotation memory. The Lorentz compression engine of the quotation memory receives the calibrated dynamic attenuation factor and reads the parameters. The parameter storage area of ​​the Lorentz compression engine is replaced with the new dynamic attenuation factor value. The new dynamic attenuation factor value recalculates the Lorentz velocity transformation parameters. After the Lorentz velocity transformation parameters are updated, new Lorentz transformation matrix parameters are generated and take effect immediately. The Lorentz compression engine applies the new Lorentz transformation matrix parameters to transform the four-dimensional spatiotemporal coordinate vector, and outputs the compressed state as the processing result, completing the real-time update.

[0130] This embodiment also provides a decision analysis system for a standard material procurement system, including: a feature acquisition module, an attenuation modeling module, a quantum compression module, a supplier screening module, an entropy reduction decision module, and a closed-loop calibration module; the feature acquisition module is used to collect historical material procurement data streams and extract monetary values, supplier credit ratings, and logistics cost combinations as basic feature sets; the attenuation modeling module is used to perform big data analysis and attenuation modeling on the timeliness attributes of the basic feature sets, establish a hyperbolic attenuation model, and output a dynamic attenuation factor; the quantum compression module is used to map the basic feature sets into coordinate vectors, input them along with the dynamic attenuation factor into a Lorentz compression engine, compress and store them as price quotes. The system comprises four modules: a supplier selection module (receiving procurement instructions, identifying material code parameters in the instructions, retrieving supplier records that meet spatiotemporal constraints from the quotation memory, and generating a candidate quotation set by sorting quotation values); an entropy reduction decision module (receiving cost parameters, budget parameters, and supply chain risk parameters submitted based on the candidate quotation set, reducing their dimensionality to a three-dimensional decision vector, using a thermodynamic entropy reduction algorithm to converge the angle of the three-dimensional decision vector, and outputting procurement plans with entropy values ​​lower than a preset entropy threshold); and a closed-loop calibration module (converting procurement plans into electronic orders and distributing them to the supplier platform, collecting performance delay data, and calibrating dynamic attenuation factors to complete the closed loop).

[0131] This embodiment also provides a computer device applicable to the decision analysis method of a standard material procurement system, comprising: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the decision analysis method of the standard material procurement system proposed in the above embodiment.

[0132] The computer device can be a terminal, comprising a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0133] This embodiment also provides a storage medium on which a computer program is stored. When executed by a processor, the program implements the decision analysis method for implementing a standard material procurement system as proposed in the above embodiments. The storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Erasable Programmable Read Only Memory (EPROM), Programmable Red-Only Memory (PROM), Read-Only Memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk.

[0134] In summary, this invention utilizes a hyperbolic decay model to dynamically model the timeliness attributes of the basic feature set. By leveraging the nonlinear coupling mechanism between market volatility and logistics cost drift factors, the dynamic decay factor autonomously adapts to changes in the market environment. The hyperbolic decay model uses supplier credit ratings as a dynamic adjustment variable, combined with industry benchmark logistics costs, to achieve precise control of the decay coefficient, thereby improving the fidelity of the timeliness value of historical price data.

[0135] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A decision analysis method for a standard material procurement system, characterized in that: include, Collect historical material procurement data streams and extract the basic feature set based on monetary values, supplier credit ratings, and logistics cost combinations; Big data analysis and attenuation modeling are performed on the time-related attributes of the basic feature set to establish a hyperbolic attenuation model and output a dynamic attenuation factor; The basic feature set is mapped to a coordinate vector, and the dynamic decay factor is input into the Lorentz compression engine for compressed storage as a quotation memory. Receive procurement instructions, identify the material code parameters in the procurement instructions, retrieve supplier records that meet the time and space constraints in the quotation memory, and generate a candidate quotation set by sorting the quotation value. Receive cost parameters, budget parameters, and supply chain risk parameters submitted based on the candidate quotation set, and convert them into a three-dimensional decision vector by reducing their dimensionality. Use the thermodynamic entropy reduction algorithm to converge the angle of the three-dimensional decision vector and output the procurement plan with an entropy value lower than the preset entropy threshold. The procurement plan is converted into electronic orders and distributed to the supplier platform. Performance delay data is collected, and the dynamic attenuation factor is calibrated to complete the closed loop.

2. The decision analysis method for the standard material procurement system as described in claim 1, characterized in that: The combination is a basic feature set, and the specific steps are as follows. The system uses a regular expression parsing engine to scan the structured fields of historical material procurement data streams, extracting the monetary value field, supplier credit rating field, and logistics cost field to generate an original feature set. Apply hyperbolic time-decay weighting to the timestamp attributes of the original feature set to generate a time-weighted feature set. The time-weighted feature set is input into the Pareto encoder for three-dimensional vector normalization mapping to generate a standard feature vector set, which is then stored as the basic feature set.

3. The decision analysis method for the standard material procurement system as described in claim 2, characterized in that: The specific steps for establishing the hyperbolic attenuation model and outputting the dynamic attenuation factor are as follows. Using the monetary value in the basic feature set as the search key, perform a matching query on the amount with a tolerance lower than the preset tolerance threshold in the historical purchase order database to obtain the order identifier; The actual transaction price is obtained by querying historical purchase fulfillment records using order identifiers. Based on the ratio of the actual transaction price to the supplier's initial quotation, an actual price validity sequence is generated. Market volatility is calculated using supplier credit ratings based on a set of basic features, and benchmark validity is estimated using logistics costs. A hyperbolic decay model is established and optimized by inputting the actual price validity sequence into a quantum annealer, outputting a dynamic decay factor.

4. The decision analysis method for the standard material procurement system as described in claim 3, characterized in that: The compressed storage is a quotation memory, and the specific steps are as follows. The monetary values, supplier credit ratings, and logistics costs of the basic feature set are normalized and mapped into coordinate vectors, and a four-dimensional spatiotemporal coordinate vector is constructed by combining the dynamic decay factor. Based on market volatility, the Lorentz transformation velocity parameters are calculated, and the four-dimensional spatiotemporal coordinate vector is spatiotemporally compressed to generate a compressed state. The compressed state is input into the quantum error correction encoder, and logical qubits are generated using the surface code algorithm. Logical qubits are written into quantum memory cells and associated with suppliers to form a quotation memory bank.

5. The decision analysis method for the standard material procurement system as described in claim 4, characterized in that: The specific steps for generating a candidate bid set by sorting bids by value are as follows. Receive procurement instructions, parse the material codes, procurement quantities, and urgency parameters in the procurement instructions, and generate demand quantum states through quantum amplitude encoding; Extract supplier quantum states from the quotation memory and construct a dynamic hypergraph network based on logistics synergy and competition intensity; The demand quantum state and dynamic hypergraph network are input into the quantum field evolution equation, and the supplier matching degree is calculated through the interaction Hamiltonian. Based on the matching degree between the quoted price and the supplier, a hyperbolic geometric sorting is performed in the Poincaré disk model to generate a set of candidate quotes.

6. The decision analysis method for the standard material procurement system as described in claim 5, characterized in that: The specific steps for the procurement scheme where the output entropy value is lower than the preset entropy threshold are as follows. The cost parameters, budget parameters, and supply chain risk parameters submitted based on the candidate bid set will be mapped into complex quantum fields to generate quantum resonant ground states; The quantum resonant ground state is input into the Chern-Simons topological field to construct the decision connection field, and the quantum vortex dynamics equation is solved to output the vortex field. The vortex field is evolved through quantum resonance tunneling conditions, and the spatial expectation value is calculated to output a three-dimensional decision vector; Calculate the entropy value of the three-dimensional decision vector. If the entropy value is lower than the preset entropy threshold, output the procurement plan. If the entropy value is higher than the preset entropy threshold, perform quantum annealing optimization and regenerate the quantum resonant ground state.

7. The decision analysis method for the standard material procurement system as described in claim 6, characterized in that: The calibration of the dynamic attenuation factor completes the closed-loop process, and the specific steps are as follows. The procurement plan is encoded into quantum state order packages and distributed to the supplier platform, and performance delay data is collected; Construct a time delay curvature model based on the collected performance delay data, and output a curvature scalar; The curvature scalar is input into the quantum phase-locked engine to generate phase gradient data, which is then used to adjust the historical dynamic decay factor and generate a calibrated dynamic decay factor. The Lorentz compression engine, which feeds back the calibrated dynamic decay factor to the quotation memory, completes the real-time update.

8. A decision analysis system for a standard material procurement system, based on the decision analysis method for a standard material procurement system according to any one of claims 1 to 7, characterized in that: It includes a feature acquisition module, an attenuation modeling module, a quantum compression module, a supplier selection module, an entropy reduction decision module, and a closed-loop calibration module; The feature acquisition module is used to collect historical material procurement data streams and extract the basic feature set based on the amount value, supplier credit rating, and logistics cost combination. The attenuation modeling module is used to perform big data analysis and attenuation modeling on the time-related attributes of the basic feature set, establish a hyperbolic attenuation model, and output a dynamic attenuation factor. The quantum compression module is used to map the basic feature set into a coordinate vector, and input the dynamic decay factor into the Lorentz compression engine for compression and storage as a quotation memory. The supplier screening module is used to receive procurement instructions, identify material code parameters in procurement instructions, retrieve supplier records that meet the time and space constraints in the quotation memory, and generate a candidate quotation set by sorting the quotation value. The entropy reduction decision module is used to receive cost parameters, budget parameters and supply chain risk parameters submitted based on the candidate quotation set, and convert them into a three-dimensional decision vector. The thermodynamic entropy reduction algorithm is used to converge the angle of the three-dimensional decision vector and output a procurement plan with an entropy value lower than a preset entropy value threshold. The closed-loop calibration module is used to convert the procurement plan into electronic orders and distribute them to the supplier platform, collect performance delay data, and calibrate the dynamic attenuation factor to complete the closed loop.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: When the processor executes the computer program, it implements the steps of the decision analysis method for the standard material procurement system according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by the processor, it implements the steps of the decision analysis method for the standard material procurement system according to any one of claims 1 to 7.

Images