Generalized coherent quantum evidence reasoning method and device based on quantum evidence theory

Abstract

The application provides a generalized coherent quantum evidence reasoning method and device based on quantum evidence theory, which comprises the following steps: obtaining multiple groups of independent to-be-combined evidence sources, generating a generalized quantum basic probability amplitude function with phase information by combining the cardinality of the proposition subset and the phase hyperparameter after characteristic normalization and classical probability assignment; encoding the function into a quantum superposition state and initializing to obtain an initial quantum state; based on the generalized quantum evidence combination rule, quantum evolution is completed in stages relying on a coherent combination quantum circuit to obtain a target quantum state; a double-layer quantum measurement mechanism is constructed to sample and obtain fusion distribution information; and data category determination is completed relying on sampling counting. The method retains evidence coherence terms through phase differentiation coding and stage-by-stage coherent evolution, and is matched with a double-layer measurement and decision mechanism, which can avoid decision bias caused by the loss of coherence terms, greatly reduce the time overhead of fusion calculation, and improve the reasoning classification accuracy and operation efficiency while meeting the quantum evidence fusion constraints.

Description

Technical Field

[0001] This application relates to the field of computer systems based on specific computational models in the next generation of information technology, and in particular to a generalized coherent quantum evidence reasoning method and apparatus based on quantum evidence theory. Background Technology

[0002] The rapid development of quantum artificial intelligence and uncertainty reasoning technologies has led to the widespread application of quantum evidence theory in scenarios such as multi-source information fusion and sample pattern classification, thanks to its advantages in open-world conflict resolution. Quantum circuits, relying on the quantum superposition property, have overcome the exponential computing power bottleneck of classical evidence combinations. Theories such as differential privacy and coherent interference are gradually being implemented, providing technical support for intelligent decision-making with complex coupled evidence.

[0003] However, traditional incoherent quantum evidence combination methods still have significant shortcomings in practical applications. First, current mainstream algorithms assume that there is no quantum coherence effect between pieces of evidence, directly discarding probability amplitude and phase information and ignoring the constructive and destructive interference laws caused by the inherent coupling of evidence. This leads to the permanent loss of coherent terms during the fusion of high-dimensional coupled data, ultimately causing significant biases in classification decisions. Second, the computational cost of traditional analytical coherent evidence combination algorithms increases exponentially, making them difficult to deploy on engineering datasets due to computational constraints. Furthermore, existing quantum coherence implementation schemes cannot simultaneously achieve computational efficiency and the ability to identify unknown samples in open worlds while preserving complete semantic information. Summary of the Invention

[0004] This application proposes a generalized coherent quantum evidence reasoning method and device based on quantum evidence theory. It aims to eliminate decision bias by fully preserving the coherent characteristics of evidence, and the dual-layer measurement mode balances accuracy and computational efficiency. It simplifies the decision-making process by eliminating normalization calculations, significantly reduces computational complexity, and can identify unknown categories in the open world, thus significantly improving the performance of quantum evidence fusion reasoning.

[0005] In a first aspect, embodiments of this application provide a generalized coherent quantum evidence reasoning method based on quantum evidence theory, including: For quantum evidence fusion scenarios containing coherent terms, multiple independent evidence sources to be combined are obtained to obtain the corresponding generalized quantum fundamental probability amplitude function. The generalized quantum fundamental probability amplitude function is a quantum state characterization function that represents the confidence of the evidence source. The generalized quantum fundamental probability amplitude function is encoded as a quantum superposition state and initialized to a pre-constructed composite quantum system to obtain the initial quantum state; Based on the initial quantum state, and based on the generalized quantum evidence combination rule, a coherent combination quantum circuit is constructed on the composite quantum system to perform a three-stage evolution to obtain the target quantum state. The three-stage evolution sequentially performs the following operations on the initial quantum state: performing intersection operation on the focal element, erasing path information and generating a coherent superposition state, and filtering and marking valid coherent superposition terms. A two-layer quantum measurement mechanism based on post-selection projection is constructed on a composite quantum system. The generalized quantum fundamental probability measurement operator is used to perform quantum measurement on the target quantum state to extract the generalized quantum fundamental probability distribution, or the generalized quantum likelihood distribution measurement operator is used to perform quantum measurement on the target quantum state to extract the generalized quantum likelihood distribution corresponding to a single proposition subset. Data category determination is performed based on the generalized quantum fundamental probability distribution or the generalized quantum likelihood distribution.

[0006] Secondly, embodiments of this application provide a generalized coherent quantum evidence reasoning device based on quantum evidence theory, comprising: The function determination unit is used to obtain multiple independent evidence sources to be combined for quantum evidence fusion scenarios containing coherent terms, so as to obtain the corresponding generalized quantum fundamental probability amplitude function. The generalized quantum fundamental probability amplitude function is a quantum state characterization function that characterizes the confidence of the evidence source. An initialization unit is used to encode the generalized quantum fundamental probability amplitude function into a quantum superposition state and initialize it to a pre-constructed composite quantum system to obtain an initial quantum state; The combinatorial evolution unit is used to construct coherent combinatorial quantum circuits on the composite quantum system based on the initial quantum state and the generalized quantum evidence combinatorial rules, so as to perform a three-stage evolution to obtain the target quantum state. The three-stage evolution sequentially performs the following operations on the initial quantum state: performing intersection operation on the focal element, erasing path information and generating a coherent superposition state, and filtering and marking valid coherent superposition terms. A quantum measurement unit is used to construct a two-layer quantum measurement mechanism based on post-selection projection on a composite quantum system. It performs quantum measurement on the target quantum state using a generalized quantum fundamental probability measurement operator to extract the generalized quantum fundamental probability distribution, or performs quantum measurement on the target quantum state using a generalized quantum likelihood distribution measurement operator to extract the generalized quantum likelihood distribution corresponding to a single proposition subset. The classification decision unit is used to determine the data category based on the generalized quantum fundamental probability distribution or the generalized quantum likelihood distribution.

[0007] Thirdly, embodiments of this application provide an electronic device including a processor, a memory, and one or more programs, the one or more programs being stored in the memory and configured to be executed by the processor, the programs including instructions for performing the steps as described in the first aspect of embodiments of this application.

[0008] Fourthly, embodiments of this application provide a computer-readable storage medium having a computer program / instructions stored thereon, which, when executed by a processor, implement the steps in the first aspect of embodiments of this application.

[0009] Fifthly, embodiments of this application provide a computer program product, including a computer program / instructions, which, when executed by a processor, implement some or all of the steps described in the first aspect of embodiments of this application.

[0010] As can be seen, in this embodiment, by using the cardinality of propositional subsets to distinguish positive and negative phases and combining it with phase hyperparameters to achieve differentiated phase encoding, the probability amplitude phase information is completely preserved, thus solving the problem of decision bias caused by the discarding of phase and loss of coherent terms in traditional incoherent algorithms from the source. Through three-stage coherent combination quantum circuit step-by-step operation, the focal element intersection operation is first completed using the computational unitary operator, then the quantum coherence is excited by erasing path information through the global Adamamen, and finally, the effective constructive coherent terms are selected and marked by the filtering operator, accurately reproducing the constructive and destructive interference laws between evidence. This scheme sets up a two-layer differentiated measurement architecture, with the confidence layer jointly measuring to reconstruct the complete probability distribution, and the likelihood layer using single-bit edge measurement to extract counts in a lightweight manner, balancing classification completeness and computational speed. The decision stage directly compares the sampled counts to complete the classification, eliminating the normalization calculation step and compressing the original exponential computational overhead to polynomial complexity. It can achieve accurate classification of closed worlds and trigger early warnings to identify unknown samples in open worlds based on empty set counts, effectively improving the practicality and inference accuracy of quantum evidence fusion algorithms. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 This is a flowchart illustrating a generalized coherent quantum evidence reasoning method based on quantum evidence theory provided in an embodiment of this application. Figure 2 This is a schematic diagram of a computational unitary operator for focal element intersection operation provided in an embodiment of this application; Figure 3 This is a schematic diagram of a filtering operator for coherent item screening provided in an embodiment of this application; Figure 4 This is a schematic diagram of a quantum evidence fusion reasoning process provided in an embodiment of this application; Figure 5 This is a functional unit block diagram of a generalized coherent quantum evidence reasoning device based on quantum evidence theory provided in an embodiment of this application; Figure 6 This is a functional unit block diagram of another generalized coherent quantum evidence reasoning device based on quantum evidence theory provided in this application embodiment; Figure 7 This is a structural block diagram of an electronic device provided in an embodiment of this application. Detailed Implementation

[0013] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0014] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0015] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0016] This application provides a generalized coherent quantum evidence reasoning method based on quantum evidence theory. The embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0017] Please see Figure 1 , Figure 1 This is a flowchart illustrating a generalized coherent quantum evidence reasoning method based on quantum evidence theory, provided in an embodiment of this application. The method includes: Step S101: For a quantum evidence fusion scenario containing coherent terms, obtain multiple independent evidence sources to be combined to obtain the corresponding generalized quantum fundamental probability amplitude function.

[0018] The generalized quantum fundamental probability amplitude function is a quantum state characterization function that represents the confidence level of the evidence source. The magnitude represents the confidence level of the evidence, and the phase carries coherent control information. It is used to realize the conversion of classical evidence into quantum encoded data, providing fundamental data for the evolution of quantum circuits.

[0019] The quantum evidence fusion scenario is an application environment for multi-source uncertain information, relying on quantum evidence theory to carry out evidence confidence fusion and classification reasoning, and is subject to quantum coherent interference effects. The evidence sources to be combined are original evidence data collected from different observation dimensions and participating in the fusion operation. The data sources of each evidence source do not interfere with each other, and their statistical characteristics are independent of each other, satisfying the operational premise of quantum evidence combination. The coherent term refers to the superposition component generated by probability amplitude phase interference, which determines the constructive / destructive interference law of the evidence and directly affects the accuracy of the fusion result.

[0020] The specific steps for obtaining multiple independent sources of evidence to be combined can be as follows: collect the original sample data to be classified; split the sample features from different observation dimensions and construct evidence separately for each dimension; ensure that the data sources of each dimension do not interfere with each other, thus obtaining multiple sets of independent sources of evidence to be combined. The scenario determines the attributes of the evidence, which means that even if the data sources of the evidence are independent of each other, the underlying features of the samples are coupled, and the evidence will still produce coherent terms after being encoded into probability amplitudes.

[0021] This embodiment elaborates in detail the generalized quantum evidence combination rule that includes coherence terms ( - The complete quantum physics implementation framework and formula derivation process are described. The entire target quantum system is divided into four functional registers: a filter flag bit of size 1 qubit ( ), an open-world indicator bit of size 1 qubit ( ), size is Closed-world data bits of qubits ), and two each containing A data register for evidence sources containing qubits. Among them, the parameters... The physical meaning of "in quantum evidence theory" is the size of the identification framework, that is, the total number of categories in the system's basic mutually exclusive propositions or pattern classification tasks; at the quantum physical implementation level, Representatives in order to utilize quantum systems A set of orthogonal ground states is used to completely isomorphically map the power set of the identification framework (containing) The number of data qubits required for a possible focal element. The range of values ​​is A positive integer. Its calculation method is as follows: That is, directly taking the identification framework (Depend on A non-empty set consisting of mutually exclusive and complete elements (i.e., all possible basic pattern categories or independent hypothetical targets) is denoted as . ） The number of independent elements in the table. For example, in a typical pattern recognition experiment, if dealing with a classification task with 3 categories, then... At this point, 3 qubits are needed for encoding. Each focal point state; if processing a 10-class classification task, then .

[0022] In one possible embodiment, for a quantum evidence fusion scenario containing coherent terms, multiple independent sources of evidence to be combined are acquired to obtain the corresponding generalized quantum fundamental probability amplitude function. This includes: for each source of evidence to be combined in the quantum evidence fusion scenario, acquiring normalized continuous feature data corresponding to the currently processed source of evidence to be combined; performing classical probability assignment on the normalized continuous feature data to generate a classical generalized fundamental probability assignment for the currently processed source of evidence to be combined; determining the phase sign based on the cardinality of the proposition subset corresponding to the classical generalized fundamental probability assignment; and adjusting the phase of the classical generalized fundamental probability assignment based on a preset phase hyperparameter to obtain a generalized quantum fundamental probability amplitude function containing phase information for the currently processed source of evidence to be combined.

[0023] In this framework, single propositional subsets correspond to positive phase signs, while multi-element composite sets within the propositional subsets correspond to negative phase signs. The propositional subsets are subsets within the identification frame. Phase hyperparameters are used to control the phase angle.

[0024] Furthermore, the identification framework consists of several mutually exclusive and complete basic category elements, and the proposition subset is specifically a set of events formed based on the combination of identification framework elements; the cardinality of the proposition subset is the total number of framework elements contained in a single proposition subset; the classical generalized basic probability assignment is used to quantify the confidence level corresponding to each proposition; the phase hyperparameter is a pre-set fixed parameter that uniformly controls the phase change amplitude.

[0025] For example, suppose there are two sources of evidence to be combined in a specific pattern classification or actual physical observation task within a quantum evidence fusion scenario. First, for each of these two sources, we obtain their corresponding Generalized Quantum Basic Probability Amplitude (GQBPA), as shown in the following expressions: ; in, The identifier GQBPA represents a quantum confidence encoding in complex form, which is the core physical quantity that is the final output of step S101 in this scheme. middle, Refers to the sources of evidence to be combined and integrated. For the source of evidence number, Each corresponds to two independent sources of evidence to be combined, and corresponds one-to-one with the first source of evidence register and the second source of evidence register in the composite quantum system. This represents the various subsets of propositions (i.e., specific pattern categories or combinations of categories) within the identification framework. The modulus represents the probability amplitude, which is a real-valued term, and the modulus represents the probability magnitude of the nth probability term. h Source of evidence, No. i The classical generalized basic probability assignment corresponding to each subset of propositions has a modulus that directly represents the confidence level of the corresponding proposition. This modulus is obtained by solving the classical probability assignment operation on the normalized continuous features. The initial phase angle represents the generalized quantum fundamental probability amplitude, and the first phase angle represents the second phase angle. h The total phase deflection angle corresponding to the i-th propositional subset is generated by two rules: the phase sign is determined based on the cardinality of the elements contained in the propositional subset, the propositional subset corresponding to a single element is assigned a positive phase sign, and the composite propositional subset containing multiple elements is assigned a negative phase sign; and the determined phase sign is multiplied by the globally unified pre-set phase hyperparameter, and the result is the complete phase angle of the corresponding propositional subset. i For the proposition subset index number, i Traverse all valid focal elements within the identification framework to distinguish different categories and subsets of composite events; index position i The imaginary unit satisfies i 2 =-1, imaginary unit i subscript of proposition index i These are two completely independent variables with no correlation. is a natural constant, the Euler complex exponent basis, used in quantum mechanics to standardize the expression of phase information.

[0026] As can be seen in this example, the features are first normalized to eliminate dimensional interference, and classical probability assignments are generated from the features. Positive and negative phases are distinguished by subset cardinality, and the phase is controlled by phase hyperparameters. Phase information is retained during the encoding stage, coherent terms are generated in a controllable manner, the confidence loss of composite propositions is reduced, and a complete quantum probability amplitude is generated, providing reliable input for subsequent quantum evolution and improving the accuracy of fusion and classification.

[0027] In one possible embodiment, the classical generalized fundamental probability assignment is phase-adjusted according to a preset phase hyperparameter to obtain a generalized quantum fundamental probability amplitude function containing phase information for the currently processed evidence source to be combined. This includes: obtaining a subset of propositions corresponding to the classical generalized fundamental probability assignment; determining whether the cardinality of the proposition subset is equal to one; when the cardinality of the proposition subset is equal to one, determining that the phase sign is positive; when the cardinality of the proposition subset is greater than one, determining that the phase sign is negative; and using the product of the phase sign and the phase hyperparameter as the phase angle, combining it with the amplitude of the classical generalized fundamental probability assignment to generate a generalized quantum fundamental probability amplitude function in complex form.

[0028] In this embodiment, based on the principle of amplitude-phase joint encoding, the original features are first normalized to smooth out the differences in data dimensions between different features. Then, the confidence level is assigned based on the processed feature values ​​to obtain the classical probability assignment representing the confidence. Subsequently, the number of elements contained in each proposition subset is counted. If the subset contains only a single element, a positive phase is configured, and if it contains multiple elements, a negative phase is configured. Finally, the phase deflection degree is uniformly constrained using a pre-set phase hyperparameter. On the basis of the original classical confidence data, exclusive phase information is superimposed to generate a generalized quantum fundamental probability amplitude that carries both confidence and coherence attributes.

[0029] The mathematical expressions for the phase coefficient, phase angle, and generalized quantum fundamental probability amplitude corresponding to this step are as follows: ; ; in, For phase sign coefficient, Representing a subset of propositions The cardinality (number of elements); This is the final phase angle; The phase hyperparameter is globally uniform; Assignment (amplitude) to the classical generalized basic probability; The generalized quantum fundamental probability amplitude in complex form; e The natural constant, the exponential position i The imaginary unit ( i 2 =-1), which is unrelated to the propositional subscript and is the standard expression for the phase of a quantum complex number.

[0030] As can be seen, in this example, by relying on the design of distinguishing phases by the number of subset elements, phase data is completely retained during the encoding stage, which is adapted to the characteristic of naturally generated coherent terms in the scenario and avoids the loss of coherent information caused by directly discarding phases in traditional processing methods. Phase fluctuations are uniformly controlled through phase hyperparameters, which delays the rapid loss of confidence in composite propositions, reduces classification errors in subsequent evidence fusion, and optimizes the accuracy of final sample classification.

[0031] Step S102: Encode the generalized quantum fundamental probability amplitude function into a quantum superposition state and initialize it to a pre-constructed composite quantum system to obtain the initial quantum state.

[0032] For example, suppose the size of the recognition frame is , ( , To identify the frame, ), For a subset of propositions The corresponding binary bit mapping vector: if identifying elements in the frame ∈ Then the bit The value is 1; if Then the bit The value is 0.

[0033] Furthermore, the quantum superposition state expression corresponding to a single source of evidence is as follows: ; ; in, For the first h The complete quantum superposition state corresponding to the combined evidence sources; To identify the power set consisting of all subsets of the frame Θ, summation is performed to traverse all propositional subsets within the frame. For the first h Source of evidence, No. j The generalized quantum fundamental probability amplitude corresponding to a subset of propositions serves as the quantum ground state. Weighting coefficients; For a subset of propositions The corresponding binary basis vectors are derived from Generated by direct product of single qubits; for Tensor direct product operation of 1 qubit; To identify the total number of basic category elements contained in the frame Θ; For the first j A subset of propositions, the first i Each identification element corresponds to a binary bit identifier, which can only take the value 0 or 1; Furthermore, the overall initial direct product state of the composite quantum system composed of multiple registers: ; in, This represents the overall initial state of the entire composite target quantum system before the application of the combined evolution circuit. Represents the filter flag (denoted as ).F The initialization state of the register, which occupies 1 qubit. This represents the initialization state of the Open-world indicator (denoted as out) register, which occupies 1 qubit. This represents the initialization state of the Closed-world data (IN) register, which is a zero-based ground state, and the register occupies... One quantum bit. and The quantum states representing the data registers of the two evidence sources each carry the generalized quantum fundamental probability assignment information corresponding to the two independent evidence sources, and each occupies a specific quantum state. One quantum bit.

[0034] Among them, quantum superposition refers to a quantum expression formed by the weighted combination of the probability amplitudes of various propositions; a composite quantum system refers to a multi-qubit quantum system composed of a filter flag register, an open-world indicator register, a closed data register, and two evidence source registers. The functions of each register are detailed as follows: Evidence source register: used to store quantum-coded data from two independent evidence sources, with the number of qubits matching the size of the identification framework. Closed data register: stores the result of focal element operation, with a number of bits. Open World Indicator Register and Filter Flag Register: Both are 1-bit dedicated registers, which respectively identify open category information and coherent filtering status. This refers to the tensor direct product operation between multiple register quantum states, representing that each register hardware independently and in parallel stores quantum information.

[0035] This step transforms the probability amplitude data into a quantum-computable form, completing the initialization of the quantum system. Following quantum mapping rules, a quantum superposition state is constructed using the probability amplitude as the ground state weight coefficient. This generated superposition state is then loaded into a pre-configured composite quantum system to generate the system's initial quantum state. This initial state is used to convert classical probability data into quantum states, providing compliant initial input for the subsequent three-stage coherent combined quantum circuit evolution.

[0036] Furthermore, this step takes the complex-type generalized quantum fundamental probability amplitude generated upstream, encodes it based on the quantum superposition property, and then loads the superposition state as a whole into a multi-register hardware architecture. The entire initialization process fully preserves the amplitude and phase information of the probability amplitude, does not destroy the previously configured coherence conditions, and ensures that subsequent quantum gates such as the Adamama gate and unitary operator can normally excite constructive / destructive interference effects.

[0037] Step S103: Based on the initial quantum state and the generalized quantum evidence combination rule, construct a coherent combination quantum circuit on the composite quantum system to perform a three-stage evolution and obtain the target quantum state.

[0038] The three-stage evolution process sequentially performs the following operations on the initial quantum state: performs intersection operations on the focal element, erases path information and generates a coherent superposition state, and filters and marks valid coherent superposition terms. A coherent combinatorial quantum circuit refers to a customized quantum computation circuit; path information refers to the intrinsic evolution path data of the quantum ground state; and valid coherent superposition terms are constructive coherent components that can be used for subsequent classification.

[0039] Furthermore, this step excites quantum coherence through a three-stage evolution, transforming the initial quantum state into a target quantum state carrying fused information. The first stage utilizes a unitary operator to perform the intersection operation of the two-evidence focal elements, generating an incoherent intermediate state. The second stage employs a global Hadamard gate (the fundamental unitary quantum logic gate for single qubits) to erase the original path information, breaking the orthogonality constraint to generate a coherent superposition state. The third stage fully preserves phase and coherence information, avoiding the loss of coherent terms, eliminating invalid noise components, and improving the accuracy of subsequent measurements. This step improves the accuracy of subsequent measurements by fully preserving phase and coherence information, avoiding the loss of coherent terms, and eliminating invalid noise components.

[0040] In one possible embodiment, the composite quantum system includes a first evidence source data register, a second evidence source data register, a closed-world data register, and an open-world indicator register, as well as a filter flag register. The first and second evidence source data registers are used to store the quantum-coded data of the evidence sources to be combined. The closed-world data register is used to store the closed-world focal element operation results. The open-world indicator register is used to identify open-world category information, and the filter flag register is used to mark the coherent screening state. Based on the initial quantum state and the generalized quantum evidence combination rules, a coherent combination quantum circuit is constructed on the composite quantum system to perform a three-stage evolution to obtain the target quantum state, including: storing the focal element data carried by the initial quantum state into... First evidence source data register and second evidence source data register; apply computational unitary operators to perform set-theoretic intersection operations on the focal data in the first and second evidence source data registers, and map the intersection results to the closed-world data register and the open-world indicator register to generate an incoherent preliminary combined intermediate state; apply an Adamama gate operation to all qubits in the first and second evidence source data registers to generate a superposition state containing coherent terms and auxiliary coherence indices; perform state flipping on the subspace where the auxiliary coherence index is zero, and physically distinguish the open-world belief terms from the conflict terms through the open-world indicator register and the closed-world data register, and mark the valid coherent superposition terms to obtain the target quantum state.

[0041] Specifically, the non-empty intersection result corresponding to the intersection result is mapped to the closed-world data register, the empty intersection result is used as the conflict term, and the open-world belief term corresponds to the empty intersection result. When the auxiliary coherence indices of the first evidence source data register and the second evidence source data register are both zero, the corresponding coherence term is equal to the coherent orthogonal sum, and the auxiliary coherence index is a binary index used to mark the superposition state components.

[0042] Among them, the computational unitary operator is a dedicated quantum transformation operator for realizing the intersection operation of the focal element set; the Adamama gate is a fundamental quantum logic gate that can break the orthogonality constraint of the quantum ground state to generate coherent components; the auxiliary coherence index is a binary code identifier used to distinguish different components within the superposition state; the conflict term refers to the invalid confidence data corresponding to the non-intersection of the focal elements of two pieces of evidence; the open-world belief term corresponds to the empty intersection of the focal elements, representing the unknown category confidence of the sample to be tested; and the five types of registers each perform their respective functions to form a complete composite quantum system. The overall function of this step is to rely on staged quantum circuit operations to complete the evidence fusion evolution within the quantum system, separate valid coherent data from invalid conflict data, and output the target quantum state that can be used for subsequent measurements.

[0043] Furthermore, this embodiment is based on the generalized quantum evidence combination rule and operates according to the three-stage quantum evolution logic. First, the focal element data in the initial quantum state is stored in two-way evidence registers. The focal element intersection operation is completed with the help of the computational unitary operator. The non-empty intersection data is stored in the closed world data register. The empty intersection content is recorded as the conflict term and the open world belief term respectively and stored in the open world indicator bit register to obtain the incoherent intermediate quantum state. Then, the Adamamen global transformation is uniformly performed on all stored bits to erase the original evolution path information and generate a coherent superposition state with a binary auxiliary coherent index. Finally, the quantum subspaces with both indices being zero are selected and the state is flipped. The open world belief term and the conflict term are separated from the open world belief term and the conflict term at the hardware level by relying on two dedicated registers. The compliant and valid coherent terms are selected and retained to finally form the target quantum state.

[0044] The first stage involves the intersection of focal elements, and this stage uses the unitary operator. Complete the intersection operation of focal set theory, with the initial quantum state denoted as . After evolution, an incoherent preliminary combination intermediate state is obtained. The expression is as follows: ; in, The computational unitary operator is a dedicated quantum transformation operator for this stage, used to implement the set-theoretic intersection operation of the focal elements of the two evidence paths; The initial quantum state of the system output by S102; This is an incoherent intermediate state obtained from the first stage of evolution; The focal element stored in the data register of the first evidence source; The focal element stored in the data register of the second evidence source; for and The non-empty intersection focal element is mapped to the closed-world data register. ; This indicates that the intersection of two focal elements is empty, corresponding to the conflict term and the open-world belief term. The result is mapped to the open-world indicator bit register. ; , This represents the quantum ground state of the two evidence source registers; , , These represent the initial zero state of the filter flag, open world indicator, and closed world data register, respectively.

[0045] Furthermore, the second stage involves generating a coherent superposition state using a global Adama gate. This stage involves the two evidence registers. , Applying a global Adama gate to all qubits By erasing path information and stimulating coherent effects, a coherent superposition state is obtained. ; ; To simplify the expression for the second intermediate state, the specific process is redefined, and the effective combination coefficient is defined. With Open World Beliefs : ; ; in, This represents the non-empty target subset generated after finding the intersection of two pieces of evidence, i.e., satisfying... and . , These represent two sources of evidence corresponding to Jiao Yuan. and The generalized quantum fundamental probability amplitude (GQBPA, i.e., complex probability amplitude). As an auxiliary coherence index, it represents the quantum computing ground state label of the vast superposition space excited by the global transformation of the Hadamard gate. They respectively represent the subsets corresponding to propositions. and A binary indicator vector (e.g., the bit is 1 if an element exists in a subset, and 0 otherwise). This is the phase factor induced during global superposition operations. The physical meaning of: represents the generation of a specific focal element. The effective combination coefficients. It strictly depends on the coherence index and incorporates constructive or destructive interference effects in the superposition of multiple bit states through phase modulation. When and When the coefficient is exactly equivalent to the coherent orthogonal sum, the coefficient is perfectly equivalent to the coherent orthogonal sum. The physical meaning: It represents that the target event is located within a known identification framework. The open-world belief beyond the empty set (i.e., the generalized quantum fundamental probability of the empty set), and the open-world belief magnitude corresponding to the empty intersection. Because the empty set... Explicitly encoded as an all-zero bit string 0, its inner product with any auxiliary index is strictly zero (i.e., ... This results in the phase factor being always equal to 1.

[0046] The derivation shows that when the auxiliary coherence index... and At this point, the phase factors completely cancel each other out, and the square of the modulus is exactly equivalent to the theoretically required combined generalized quantum fundamental probability distribution. The second intermediate state can then be rewritten as: .

[0047] Finally, the third stage involves generating the target quantum state using a filtering operator. This stage employs a filtering operator. right , The target subspace is flipped to distinguish between open-world belief terms and conflict terms, ultimately yielding the target quantum state. : ; in, The unitary transformation operator is used as a filtering operator to perform subspace state reversal, coherence term filtering, and labeling. The ground state after the filter flag register is flipped is used to mark valid coherent terms; the first term is the valid coherent term corresponding to the subspace, which is marked as usable fused data; the second term is the open world belief term corresponding to the empty intersection, which is separately identified by the open world indicator register; the third term is the invalid coherent noise component with a non-zero index, which is not marked as valid.

[0048] As can be seen, in this example, the quantum coherent information is fully preserved through hierarchical circuit evolution, overcoming the computational bias caused by the direct discarding of coherent terms in traditional incoherent algorithms; and the physical isolation of open data and conflicting data is achieved by relying on register hardware partitioning, which simplifies the subsequent data extraction process and optimizes the accuracy of subsequent sampling measurement and category determination.

[0049] In one possible embodiment, the process of state flipping the subspace with an auxiliary coherence index of zero, and physically distinguishing open-world belief terms from conflict terms using an open-world indicator register and a closed-world data register, and marking valid coherent superposition terms to obtain the target quantum state, includes: obtaining a superposition state containing coherent terms and an auxiliary coherence index; performing state flipping on the subspace with an auxiliary coherence index of zero using a filtering operator; identifying the state of the open-world indicator register in the superposition component after state flipping; identifying the state of the closed-world data register in the superposition component after state flipping; distinguishing open-world belief terms and conflict terms into different quantum state components based on the states of the open-world indicator register and the closed-world data register; and marking constructive coherent terms containing correctly fused information using a filtering operator to obtain the target quantum state.

[0050] Among them, the filtering operator is a unitary transformation operator specifically used for quantum space transformation and coherence term selection; the auxiliary coherence index is a binary identifier used to distinguish each quantum component in the superposition state; the open world belief term corresponds to the unknown category confidence data when the evidence focal elements have no intersection; the conflict term is the invalid confidence component caused by the contradiction of evidence; the constructive coherence term is the effective coherence component that can participate in the subsequent classification calculation.

[0051] In this embodiment, the indexed coherent superposition state generated in the previous step is read first. Using filtering operators Locking double indexes to all zeros (i.e.) , The target subspace is then established and quantum state flipping is performed. Next, the storage level states of the open-world indicator bit and the closed-world data register are read, and based on the storage differences between the two types of registers, the open-world belief term and conflict term are separated at the quantum level. Finally, a filtering operator is used again to lock in constructive coherent terms with effective fusion value and mark them, eliminating useless coherent noise to complete the target quantum state. The construction of.

[0052] As can be seen, in this example, the filtering operator is used to complete the subspace screening and coherent labeling, accurately retain the effective coherent information, and the register state is used to realize the physical division of the two types of data, reduce the interference of conflicting data on subsequent measurements, and improve the accuracy of subsequent sampling statistics and sample classification results.

[0053] For example, please refer to Figure 2 and Figure 3 , Figure 2 This is a schematic diagram of a computational unitary operator for focal element intersection operations provided in an embodiment of this application. Figure 3 This is a schematic diagram of a filtering operator for coherent item screening provided in an embodiment of this application. For example... Figure 2 and Figure 3 The underlying quantum hardware wiring structures of the two core dedicated unitary operators in the evolution stage of step S103 are shown respectively. Both types of operators are composed of single-qubit gates, controlled two-qubit gates, and multi-layered graded processing sub-modules connected in series: Figure 2 The dedicated computational unitary operator for the intersection operation of focal elements in the first stage. Internal circuitry: Two evidence registers are arranged in parallel hardware. , Closed data register Open indicator register Auxiliary control register The circuit incorporates a single-qubit X quantum gate and multiple sets of controlled non-CNOT gates as basic computational units, with hierarchical configuration of basic processing modules. Dimensional adaptation module Conjugate Transformation Module Conflict Differentiation Module This hardware circuit fully implements the intersection operation of two-way evidence focal element set theory, automatically mapping the non-empty intersection to... Mapping empty intersection conflict terms to It outputs an incoherent intermediate quantum state.

[0054] Figure 3 A dedicated filter operator for coherent screening in the third stage. Internal circuitry: Two evidence registers are arranged in parallel hardware. , Two auxiliary control registers , Filter flag special register Similarly, single-bit X-gates and controlled non-CNOT gates are configured, with the basic processing modules configured in a hierarchical manner. Dimensional adaptation module This hardware circuitry achieves quantum state flipping in the target subspace with an all-zero auxiliary coherent index, relying on... The register completes the marking of effective constructive coherence terms, separates the open-world belief term from the invalid conflict noise component at the hardware level, and outputs the target quantum state that can be measured.

[0055] in, and All are customized fixed quantum circuits, requiring no external software iterative control. They rely on hardware gate circuits to directly complete quantum transformations, ensuring the parallel acceleration characteristics of the three-stage evolution operation.

[0056] Step S104: Construct a two-layer quantum measurement mechanism based on post-selection projection on the composite quantum system. Perform quantum measurement on the target quantum state using the generalized quantum fundamental probability measurement operator to extract the generalized quantum fundamental probability distribution, or perform quantum measurement on the target quantum state using the generalized quantum likelihood distribution measurement operator to extract the generalized quantum likelihood distribution corresponding to the single proposition subset.

[0057] The post-selection projection is a projection transformation operation that pre-defines the effective subspace and eliminates invalid quantum components before quantum measurement. The two-layer quantum measurement mechanism consists of a hierarchical and step-by-step measurement structure. The first layer uses post-selection to filter effective quantum state components, and the second layer statistically analyzes the collapse results of the remaining effective components. The fused distribution information is the statistical result of the confidence probability corresponding to each proposition after evidence fusion. The purpose of this step is to perform hierarchical quantum measurement on the evolved target quantum state, transforming the quantum superposition state into classical statistical distribution data that can be used for classification and determination. The generalized quantum fundamental probability distribution and the generalized quantum likelihood distribution can also be collectively referred to as fused distribution information.

[0058] In this embodiment, a two-layer measurement architecture is first constructed based on the labeling information of each register in the composite quantum system. The first layer uses a post-selection projection operator for pre-screening, discarding conflict-type invalid quantum components based on the previously filtered flags, retaining only the labeled effective coherent subspace. Then, the second layer of conventional quantum projection measurement is initiated, causing quantum collapse of the filtered effective quantum states. The measurement frequency of each ground state is counted, and the confidence ratio corresponding to each proposition is obtained based on the frequency conversion. Finally, complete fused distribution information is formed. This step of using post-selection pre-screening followed by layered measurement avoids interference from invalid conflict terms with the measurement statistics, improves the statistical reliability of the fused distribution data, and provides accurate data basis for subsequent sample category determination based on the distribution information.

[0059] In one possible embodiment, a two-layer quantum measurement mechanism based on post-selection projection is constructed on a composite quantum system. This mechanism extracts the generalized quantum fundamental probability distribution by performing quantum measurement on the target quantum state using a generalized quantum fundamental probability measurement operator, or extracts the generalized quantum likelihood distribution corresponding to a single proposition subset by performing quantum measurement on the target quantum state using a generalized quantum likelihood distribution measurement operator. This includes: performing projection measurement on a filter flag register using a post-selection projection operator to eliminate noise samples that do not satisfy the zero-frequency coherence condition, obtaining an effective sampling ensemble; on the effective sampling ensemble, selecting to perform either a confidence layer measurement or a likelihood layer measurement; if a confidence layer measurement is performed, then a joint measurement is performed on the closed-world data register and the open-world indicator register using a generalized quantum fundamental probability measurement operator to obtain the sampling count corresponding to the proposition subset, which serves as the generalized quantum fundamental probability distribution; if a likelihood layer measurement is performed, then an edge measurement is performed on a single bit of the closed-world data register using a generalized quantum likelihood distribution measurement operator to obtain the edge excitation count corresponding to the single proposition subset, which serves as the generalized quantum likelihood distribution corresponding to the single proposition subset.

[0060] Among them, the post-selection projection operator is a projection transformation operator for quantum space screening; the zero-frequency coherence condition is a criterion used to distinguish between valid coherent terms and invalid noise components; the valid sampling ensemble is the set of compliant quantum samples retained after removing noise samples; the confidence layer measurement is a register joint measurement method for all proposition subsets; the likelihood layer measurement is a single-bit edge measurement method for single-element basic propositions; and the fusion distribution information relies on measurement counts to quantify the fusion confidence level of each proposition.

[0061] In this embodiment, a screening-first, layered measurement logic is adopted. First, the post-selection projection operator is used to perform projection screening on the filter flag register to remove invalid noise samples that do not meet the zero-frequency coherence condition, thus purifying the valid sampling ensemble. Subsequently, the two measurement modes are flexibly switched according to the actual reasoning needs. When the confidence layer measurement is selected, the quantum states of both closed-world and open-world registers are read simultaneously, and the sampling frequency of the full-dimensional proposition subset is statistically analyzed through joint measurement. When the likelihood layer measurement is selected, only the single-bit edge measurement is performed on the closed-world data register, and the excitation count of the basic single proposition subset is counted separately. The count values ​​obtained from both types of measurement can be converted into fused distribution information.

[0062] For example, the construction of a two-layer quantum measurement operator is to strip away coherent noise and extract the effective fusion distribution, and the definition includes a post-selection projection operator, a generalized quantum fundamental probability measurement operator used in the confidence layer measurement, and a generalized quantum likelihood distribution measurement operator used in the likelihood layer measurement.

[0063] The calculation formula for the post-selection projection operator is as follows: ; in, The subsequent selection projection operator is used to filter the effective coherent subspace by pre-filtering the flag register; Retrieve the valid flag ground state for the filter flag register. The projection operator; This is a joint unit operator for all remaining registers (open world indicator bits, closed data, and two evidence source registers), and the states of the other registers are not changed. This represents the total number of valid samples after subsequent projection filtering. This operator only retains samples corresponding to... The effective sampling, and the total sampling count obtained by this operator, is denoted as [missing information]. Furthermore, the generalized quantum fundamental probability measurement operator: ; , ; ; in, For the generalized quantum fundamental probability measurement operator corresponding to the open-world empty set in the belief layer measurement, corresponding to the unknown category belief term; For the generalized quantum fundamental probability measurement operator corresponding to the subset of non-empty propositions of the closed world in the confidence layer measurement, traverse all non-empty focal elements of the identification framework; A dedicated measurement operator for conflict terms, statistically valid invalid confidence sampling with no overlap in two pieces of evidence; For the open-world indicator bit register, there are two types of flag ground states; The non-empty focal element and empty set of the closed-world data register correspond to the ground state; Sampling and counting for unknown categories in open worlds; Sample and count for each closed subset of propositions; Invalid sample counts for conflicting items; Furthermore, the generalized quantum likelihood distribution measurement operator is used to define the action on a closed-world data register. No. k edge operator of 1 bit : ; ; The sampling counts obtained by each operator are denoted as follows: , .

[0064] in, , An edge measurement operator for distinguishing conflict and open unknown categories in the likelihood layer; For the closed-world data register unit operator, only edge observation is performed on a single bit, while the remaining bits remain unchanged; The identification framework includes the total number of bits for the basic categories; The bit index corresponding to the single proposition subset; Count the unknown categories in the likelihood layer empty set; Count the edge excitations of single-element basic propositions.

[0065] As can be seen, in this example, the post-selection pre-filter eliminates the interference of invalid noise on the measurement results. At the same time, the dual-layer measurement mode can be switched as needed. It can obtain the complete fusion reliability of the whole proposition and quickly obtain the statistical data of single-category propositions. It adapts to the classification calculation needs of both closed and open scenarios, and improves the flexibility and accuracy of the measurement results.

[0066] Step S105: Perform data category determination based on the generalized quantum fundamental probability distribution or the generalized quantum likelihood distribution.

[0067] This step involves determining the data category by obtaining the fusion distribution information of the corresponding evidence sources to be combined.

[0068] In one possible embodiment, data category determination is performed based on the generalized quantum fundamental probability distribution or the generalized quantum likelihood distribution, including: determining the subset of propositions corresponding to the maximum value in the sampled count as the classification result based on the generalized quantum fundamental probability distribution; or determining the subset of single propositions corresponding to the maximum value in the edge excitation count as the classification result based on the generalized quantum likelihood distribution; when the maximum value corresponds to an empty set, an open world warning is triggered and the data is determined to be an unknown new category.

[0069] Among them, the empty set represents the target category that does not exist within the identification framework, corresponding to samples that do not match the known classification; the open world warning is a judgment label used to mark unfamiliar and unknown samples; the sampling count is the statistical value of the whole proposition subset obtained by the confidence layer; and the edge excitation count is the statistical value of the single proposition subset obtained by the likelihood layer.

[0070] In this embodiment, the corresponding judgment rules are matched based on the data types of the two different measurement outputs mentioned above. For the full proposition sampling count output by the confidence layer, the subset of propositions with the highest count value is selected as the category to which the sample belongs. For the single-element edge excitation count output by the likelihood layer, the basic single-element category corresponding to the maximum count value is selected. If the maximum statistical value falls on the empty set component, it means that none of the existing known categories can match the current sample, and the sample is then determined to be an unknown new category in an open environment.

[0071] The formula for calculating the total number of valid non-collision samples in the system is as follows: ; In Derivation 1, the distribution reconstruction and normalized fusion confidence distribution are derived based on the quantum measurement axiom, where the physical statistical sampling frequency is equivalent to the normalized fusion confidence result. ; ; ; in, The total number of valid sample ensembles after eliminating conflicting samples; Generalized confidence of the closed-world proposition subset normalized after fusion of two pieces of evidence; Normalized confidence for the open-world empty set (unknown category) after fusion of two pieces of evidence; To assist coherent indexing , Time j The coherence combination coefficients of a proposition; The amplitude of the open-world belief term corresponding to the empty intersection; Normalize the likelihood confidence for single-element fundamental propositions; Normalized confidence for unknown categories in the likelihood layer; To identify the first within the framework k A single-element basic category.

[0072] Derivation 2 simplifies decision-making by eliminating the need for normalization logic. In the decision-making phase, the goal is to determine the maximum belief or likelihood. This is due to the normalization factor... It is a globally common constant and does not change the sorting of count sizes. Therefore, the decision operator... and Normalization calculations can be eliminated, and classification can be determined directly through the original sample counts. This is obtained by selecting a subset of single propositions corresponding to the maximum value in the reliability measurement sample count. It is obtained by selecting a subset of single propositions corresponding to the maximum value in the likelihood measurement sample count. The specific formula is: ; ; in, Output results for the global classification decision of the confidence layer; Output the results for single-element classification decisions in the likelihood layer; This operation, which takes the maximum value of the function corresponding to the independent variable, is used to locate the proposition with the highest count. It is the union of all known basic categories and the empty set (unknown category).

[0073] As can be seen, in this example, the two data formats adapted to the dual-layer measurement output are set with differentiated judgment logic, which takes into account both the full subset confidence judgment and the fine single-class judgment capability. It uses empty set recognition to realize the detection of unknown samples in the open world, overcoming the limitation of traditional evidence algorithms that cannot identify new categories.

[0074] For example, please refer to Figure 4 , Figure 4 This is a schematic diagram of a quantum evidence fusion reasoning process provided in an embodiment of this application. Figure 4 As shown, Figure 4 This application fully demonstrates a five-stage complete process from raw sample input to category determination, executed sequentially by five major functional modules: Module 0 is the raw data input module. Taking the iris flower pattern classification task as an engineering example, the input sample consists of four continuous features: sepal length, sepal width, petal length, and petal width. These features are then processed by the amplitude coding unit. Complete feature normalization and amplitude precoding, and output the original feature data of two independent evidence sources; Module 1 is the quantum input encoding module. Based on the amplitude-phase joint encoding rule mentioned above, it generates the generalized quantum fundamental probability amplitude, completes the quantum superposition state construction according to the summation formula shown in the figure, and obtains the quantum ground state stored in the two evidence registers. , Summation traversal identification framework power set The entire set of propositions within; Module 2 is the core coherent generalized quantum evidence combination circuit of this application, corresponding to the hardware carrier of the three-stage evolution in step S103. The composite quantum system includes five parallel quantum registers: two... Bit Evidence Source Register, Closed World Data Register Open World Indicator Bit Register 1-bit filter flag register Unitary operators are sequentially deployed within the circuit. Global Adama Gate H, Filtering Operator After the evolution is complete, the projection operator is selected. Complete the pre-selection process; the filter flags are clearly distinguished in the diagram. Invalid noise subspace The effective coherent subspace is pre-filtered and then connected to the unified measurement operator. ; Module 3 is the measurement and reconstruction module, corresponding to the two-layer quantum measurement mechanism in step S104, which is divided into the confidence layer measurement submodule SI and the likelihood layer measurement submodule SII: SI adopts the confidence layer measurement operator. Statistical conflict sampling number Counting of empty set unknown classes Counting of each proposition subset Based on the total number of valid samples Normalization yields fusion confidence , SII employs the likelihood layer measurement operator. Statistical counting of single propositional subset edges Empty set counting Normalized output single-class likelihood confidence , ; Module 4 is the decision-making process module, corresponding to the category determination logic in step S105. It configures the confidence layer decision operator DBP and the likelihood layer decision operator DPl, respectively. The operation selects the proposition corresponding to the maximum count as the output result; the accompanying pie chart visually demonstrates the classification distribution effect of the three known categories of irises, Setosa, Versicolour, and Virginica, as well as the unknown open categories of "other" outside the identification framework.

[0075] visible, Figure 4 The entire technical process from S101 to S105 is fully connected, and the mathematical formulas and hardware circuit structure of feature encoding, quantum evolution, two-layer measurement, and classification decision are visualized, intuitively demonstrating the core advantages of this solution in retaining quantum coherence terms throughout the entire process and being compatible with closed / open world classification.

[0076] As can be seen, in the method flow of this application, starting from the generation of quantum probability amplitude through evidence source encoding, through quantum system initialization, three-stage coherent circuit evolution, two-layer post-selection quantum measurement, and finally to the final category determination, controllable coherent information is retained in the quantum domain throughout the process. Through staged hardware register partitioning, layered measurement, and differentiated determination, it takes into account both conventional classification in the closed world and new category detection in the open world, and fully realizes quantum evidence fusion with coherent terms and intelligent sample classification.

[0077] In one possible embodiment, this application includes engineering implementation and performance verification examples. This embodiment combines publicly available machine learning datasets to conduct engineering implementation tests on the method and quantum circuit of this application, completing verification in both closed-world and open-world application scenarios, and simultaneously conducting comparative experiments on algorithm complexity and classification accuracy.

[0078] The experimental steps include data preprocessing and generation of generalized quantum fundamental probability amplitude, configuration of quantum circuit operation, classification and judgment practical rules, comparison of algorithm complexity, experimental results and technical effects.

[0079] Specifically, the first step involves data preprocessing and the generation of generalized quantum fundamental probability amplitudes. Four classic datasets—iris, wine, breast cancer, and Parkinson's disease—are selected, with continuous features used as the original input data. First, a min-max scaling algorithm is employed to scale all continuous features to the [0,1] interval, eliminating dimensional differences between features. Then, based on distance metrics, fuzzy membership rules, and other principles, classic generalized fundamental probability assignments are generated from the normalized features. .

[0080] In this process, the generalized quantum fundamental probability amplitude is generated according to the amplitude-phase joint encoding rule mentioned above: the phase sign is distinguished according to the cardinality of the proposition subset, the single proposition subset is configured with a positive phase, and the multi-element composite proposition is configured with a negative phase. The phase angle is uniformly controlled by the preset phase hyperparameter, and finally the complex form of GQBPA is generated, thus completing the conversion from classical data to quantum probability amplitude.

[0081] The second step is the configuration of the quantum circuit operation. The generated generalized quantum fundamental probability amplitude is encoded as a quantum superposition state and loaded into a composite quantum system composed of multiple registers; the number of quantum samplings is uniformly set. The coherent combined quantum circuit is initiated to complete a three-stage evolution, followed by a selective projection two-layer measurement. The measurement mode can be switched to joint measurement at the confidence layer or edge measurement at the likelihood layer as needed.

[0082] The third step involves practical rules for classification. This includes closed-world and open-world scenarios. In closed-world scenarios, all samples belong to known categories within the identification framework. The classification result is determined by comparing the sample count and edge excitation count; the proposition corresponding to the maximum value is the classification result. In open-world scenarios, some original categories are removed as unknown interference terms. If the measured maximum value corresponds to an empty set... This immediately triggers an open-world alert, determining that the sample is an unknown new category outside the identification framework.

[0083] Step 4: Algorithm complexity comparison. The computational complexity of the traditional coherent generalized evidence combination algorithm is... This involves exponential complexity and suffers from the severe curse of dimensionality; in the scheme of this application, the algorithm complexity corresponding to the confidence layer measurement is... The algorithm complexity for likelihood layer measurement is... Both models achieve a leap from exponential to polynomial speeds, significantly reducing computational overhead. Specifically: To identify the number of basic categories in the framework, This represents the total number of propositions related to the concept of "Jiao Yuan".

[0084] Step 5: Experimental Results and Technical Effectiveness. This experiment employed a 5-fold repeated cross-validation strategy and introduced a prediction difference rate. The consistency of the quantification algorithm is expressed by the following formula: ; in, For the entire test sample set, , These are the classification results of the two algorithms, with the index values ​​ranging from [0,1]. The closer the value is to 0, the higher the consistency of the algorithms.

[0085] Experiments show that as the number of quantum samplings gradually increases, the classification accuracy of this scheme continues to rise, and the prediction difference rate gradually approaches 0, which is highly consistent with the theoretical coherence rule. Compared with the traditional incoherent quantum evidence algorithm, this scheme fully preserves the quantum coherence term, and the classification accuracy is significantly improved in high-dimensional feature coupling scenarios. In open-world scenarios, this scheme can accurately identify unknown categories, making up for the shortcomings of traditional algorithms in detecting new samples.

[0086] As can be seen, this embodiment completes the entire process based on a real dataset, verifying the feasibility of the quantum circuit, measurement mechanism and judgment rules of this application; at the same time, through comparative experiments, it is demonstrated that this solution has low computational complexity, high classification accuracy and open scenario adaptability, and has outstanding engineering application value.

[0087] The following are embodiments of the apparatus of this application. These embodiments of the apparatus and the embodiments of the method of this application belong to the same concept and are used to execute the methods described in the embodiments of this application. For ease of explanation, only the parts related to the apparatus embodiments of this application are shown in the embodiments of this application. For specific technical details not disclosed, please refer to the description of the embodiments of the method of this application, which will not be repeated here.

[0088] This application provides a generalized coherent quantum evidence reasoning device based on quantum evidence theory. Specifically, the generalized coherent quantum evidence reasoning device based on quantum evidence theory is used to execute the steps performed by the controller in the above-described generalized coherent quantum evidence reasoning method based on quantum evidence theory. The generalized coherent quantum evidence reasoning device based on quantum evidence theory in this application may include modules corresponding to the respective steps.

[0089] This application embodiment can divide the generalized coherent quantum evidence reasoning device based on quantum evidence theory into functional modules according to the above method example. For example, each function can be divided into a separate functional module, or two or more functions can be integrated into one processing module. The integrated module can be implemented in hardware or as a software functional module. The module division in this application embodiment is illustrative and only represents one logical functional division; other division methods may be used in actual implementation.

[0090] When dividing each function into modules according to its corresponding function. Figure 5This application provides a functional block diagram of a generalized coherent quantum evidence reasoning device based on quantum evidence theory. The device 50 includes: a function determination unit 501, used to acquire multiple independent evidence sources to be combined in a quantum evidence fusion scenario containing coherent terms, to obtain corresponding generalized quantum fundamental probability amplitude functions, where the generalized quantum fundamental probability amplitude function is a quantum state characterization function representing the confidence level of the evidence source; an initialization unit 502, used to encode the generalized quantum fundamental probability amplitude function into a quantum superposition state and initialize it to a pre-constructed composite quantum system to obtain an initial quantum state; and a combination evolution unit 503, used to, based on the initial quantum state and the generalized quantum evidence combination rules, in the complex... A coherent combinatorial quantum circuit is constructed on the composite quantum system to perform a three-stage evolution to obtain the target quantum state. The three-stage evolution sequentially performs the following operations on the initial quantum state: performing an intersection operation on the focal element, erasing path information and generating a coherent superposition state, and filtering and marking valid coherent superposition terms. A quantum measurement unit 504 is used to construct a two-layer quantum measurement mechanism based on post-selection projection on the composite quantum system. It performs quantum measurement on the target quantum state using a generalized quantum fundamental probability measurement operator to extract the generalized quantum fundamental probability distribution, or performs quantum measurement on the target quantum state using a generalized quantum likelihood distribution measurement operator to extract the generalized quantum likelihood distribution corresponding to a single proposition subset. A classification decision unit 505 is used to perform data category determination based on the generalized quantum fundamental probability distribution or the generalized quantum likelihood distribution.

[0091] In one possible embodiment, in a quantum evidence fusion scenario containing coherent terms, to obtain multiple independent sources of evidence to be combined to obtain the corresponding generalized quantum fundamental probability amplitude function, the function determination unit 501 is specifically used for: for each source of evidence to be combined in the quantum evidence fusion scenario, obtaining normalized continuous feature data corresponding to the currently processed source of evidence to be combined; performing classical probability assignment on the normalized continuous feature data to generate a classical generalized fundamental probability assignment for the currently processed source of evidence to be combined; determining the phase sign according to the cardinality of the proposition subset corresponding to the classical generalized fundamental probability assignment, wherein a single proposition subset in the proposition subset corresponds to a positive phase sign, and a multi-element composite set in the proposition subset corresponds to a negative phase sign, and the proposition subset is a subset in the identification framework; adjusting the phase of the classical generalized fundamental probability assignment according to a preset phase hyperparameter to obtain the generalized quantum fundamental probability amplitude function containing phase information of the currently processed source of evidence to be combined, wherein the phase hyperparameter is used to control the phase angle.

[0092] In one possible embodiment, in adjusting the phase of the classical generalized fundamental probability assignment according to a preset phase hyperparameter to obtain the generalized quantum fundamental probability amplitude function containing phase information of the currently processed evidence source to be combined, the initialization unit 502 is specifically used to: obtain the proposition subset corresponding to the classical generalized fundamental probability assignment; determine whether the cardinality of the proposition subset is equal to one; when the cardinality of the proposition subset is equal to one, determine that the phase sign is positive; when the cardinality of the proposition subset is greater than one, determine that the phase sign is negative; and use the product of the phase sign and the phase hyperparameter as the phase angle, combine it with the amplitude of the classical generalized fundamental probability assignment, and generate a generalized quantum fundamental probability amplitude function in complex form.

[0093] In one possible embodiment, the composite quantum system includes a first evidence source data register, a second evidence source data register, a closed-world data register, an open-world indicator register, and a filter flag register. The first and second evidence source data registers are used to store the quantum-coded data of the evidence sources to be combined. The closed-world data register is used to store the closed-world focal element operation results. The open-world indicator register is used to identify open-world category information. The filter flag register is used to mark the coherent screening state. Regarding the construction of coherent combination quantum circuits on the composite quantum system based on the initial quantum state and the generalized quantum evidence combination rules to perform a three-stage evolution to obtain the target quantum state, the initialization unit 502 is specifically used to: store the focal element data carried by the initial quantum state into the first and second evidence source data registers; apply a computational unitary operator to the focal element data in the first and second evidence source data registers. The process involves performing set-theoretic intersection operations and mapping the intersection results to a closed-world data register and an open-world indicator register to generate an incoherent preliminary combined intermediate state. The non-empty intersection result is mapped to the closed-world data register, the empty intersection result is used as a conflict term, and the open-world belief term corresponds to the empty intersection result. An Adamama gate operation is applied to all qubits in the first and second evidence source data registers to generate a superposition state containing coherent terms and auxiliary coherence indices. When the auxiliary coherence indices of both the first and second evidence source data registers are zero, the corresponding coherent term equals the coherent orthogonal sum. The auxiliary coherence index is a binary index used to label the superposition state components. The subspace with zero auxiliary coherence indices is state-flipped, and the open-world belief term and conflict term are physically distinguished using the open-world indicator register and the closed-world data register. Valid coherent superposition terms are then labeled to obtain the target quantum state.

[0094] In one possible embodiment, regarding the state flipping of the subspace with an auxiliary coherence index of zero, and the physical differentiation of open-world belief terms and conflict terms through open-world indicator bit registers and closed-world data registers, and the marking of valid coherent superposition terms to obtain the target quantum state, the combinatorial evolution unit 503 is specifically used to: obtain a superposition state containing coherent terms and an auxiliary coherence index; perform state flipping on the subspace with an auxiliary coherence index of zero through a filtering operator; identify the state of the open-world indicator bit register in the superposition state component after state flipping; identify the state of the closed-world data register in the superposition state component after state flipping; differentiate the open-world belief terms and conflict terms into different quantum state components according to the state of the open-world indicator bit register and the state of the closed-world data register; and mark the constructive coherent terms containing correct fusion information through a filtering operator to obtain the target quantum state.

[0095] In one possible embodiment, a two-layer quantum measurement mechanism based on post-selection projection is constructed on the composite quantum system. This mechanism extracts the generalized quantum fundamental probability distribution by performing quantum measurement on the target quantum state using a generalized quantum fundamental probability measurement operator, or extracts the generalized quantum likelihood distribution corresponding to a single proposition subset by performing quantum measurement on the target quantum state using a generalized quantum likelihood distribution measurement operator. Specifically, the quantum measurement unit 504 is used to: perform projection measurement on the filter flag register using the post-selection projection operator to eliminate noise samples that do not meet the zero-frequency coherence condition, obtaining an effective sampling ensemble; on the effective sampling ensemble, select to perform either a confidence layer measurement or a likelihood layer measurement; if a confidence layer measurement is performed, then a joint measurement is performed on the closed-world data register and the open-world indicator register using the generalized quantum fundamental probability measurement operator to obtain the sampling count corresponding to the proposition subset, which serves as the generalized quantum fundamental probability distribution; if a likelihood layer measurement is performed, then an edge measurement is performed on a single bit of the closed-world data register using the generalized quantum likelihood distribution measurement operator to obtain the edge excitation count corresponding to the single proposition subset, which serves as the generalized quantum likelihood distribution corresponding to the single proposition subset.

[0096] In one possible embodiment, in performing data category determination based on the generalized quantum fundamental probability distribution or the generalized quantum likelihood distribution, the classification decision unit 505 is specifically used to: determine the subset of propositions corresponding to the maximum value in the sampled count as the classification result based on the generalized quantum fundamental probability distribution; or, determine the subset of single propositions corresponding to the maximum value in the edge excitation count as the classification result based on the generalized quantum likelihood distribution; when the maximum value corresponds to an empty set, an open world warning is triggered, and the value is determined to be an unknown new category.

[0097] When using integrated units, such as Figure 6 As shown, Figure 6This is a functional unit block diagram of another generalized coherent quantum evidence reasoning device based on quantum evidence theory provided in this application embodiment. Figure 6 The generalized coherent quantum evidence reasoning device 50 based on quantum evidence theory includes a processing module 602 and a communication module 601. The processing module 602 controls and manages the operations of the generalized coherent quantum evidence reasoning device 50, such as the steps of the function determination unit 501, initialization unit 502, combinatorial evolution unit 503, quantum measurement unit 504, and classification decision unit 505, and / or other processes for executing the techniques described herein. The communication module 601 supports interaction between the generalized coherent quantum evidence reasoning device and other devices. Figure 6 As shown, the generalized coherent quantum evidence reasoning device based on quantum evidence theory may include a storage module 603, which is used to store the program code and data of the generalized coherent quantum evidence reasoning device based on quantum evidence theory.

[0098] The processing module 602 may be a processor or processing module, such as a central processing unit (CPU), a general-purpose processor, a digital signal processor (DSP), an ASIC, an FPGA, or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. The processor may also be a combination that implements computational functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc. The communication module 601 may be a transceiver, RF circuitry, or a communication interface, etc. The storage module 603 may be a memory.

[0099] All relevant content in each scenario involved in the above method embodiments can be referenced from the functional descriptions of the corresponding functional modules, and will not be repeated here. The above-mentioned generalized coherent quantum evidence reasoning device 50 based on quantum evidence theory can execute the above-mentioned... Figure 1 The method for generalized coherent quantum evidence reasoning based on quantum evidence theory is shown.

[0100] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions according to the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means. A computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. Semiconductor media can be solid-state drives.

[0101] Figure 7 This is a structural block diagram of an electronic device provided in an embodiment of this application. For example... Figure 7 As shown, the electronic device 70 may include one or more of the following components: a processor 701 and a memory 702 coupled to the processor 701, wherein the memory 702 may store one or more computer programs 703, and the one or more computer programs 703 may be configured to implement the methods described in the above embodiments when executed by one or more processors 701.

[0102] Processor 701 may include one or more processing cores. Processor 701 connects to various parts within the electronic device 70 using various interfaces and lines, and performs various functions and processes data of the electronic device 70 by running or executing instructions, programs, code sets, or instruction sets stored in memory 702, and by calling data stored in memory 702. Optionally, processor 701 may be implemented using at least one hardware form of Digital Signal Processing (DSP), Field-Programmable Gate Array (FPGA), or Programmable Logic Array (PLA). Processor 701 may integrate one or more of the following: Central Processing Unit (CPU), Graphics Processing Unit (GPU), and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the displayed content; and the modem handles wireless communication. It is understood that the modem may also not be integrated into processor 701 and may be implemented separately using a communication chip.

[0103] The memory 702 may include random access memory (RAM) or read-only memory (ROM). The memory 702 can be used to store instructions, programs, code, code sets, or instruction sets. The memory 702 may include a program storage area and a data storage area. The program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as touch functionality, sound playback functionality, image playback functionality, etc.), and instructions for implementing the various method embodiments described above. The data storage area may also store data created by the electronic device 70 during use.

[0104] It is understood that the electronic device 70 may include more or fewer structural elements than those shown in the above block diagram, and no limitation is made herein.

[0105] This application also provides a computer storage medium storing a computer program / instructions thereon, which, when executed by a processor, implements some or all of the steps of any of the methods described in the above method embodiments.

[0106] This application also provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps of any of the methods described in the above method embodiments.

[0107] It should be understood that in the various embodiments of this application, the sequence number of each process does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0108] In the several embodiments provided in this application, it should be understood that the disclosed methods, apparatuses, and systems can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for example, the division of units is merely a logical functional division, and there may be other division methods in actual implementation; for example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0109] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0110] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can be physically comprised separately, or two or more units can be integrated into one unit. The integrated unit described above can be implemented in hardware or in the form of hardware plus software functional units.

[0111] The integrated units implemented as software functional units described above can be stored in a computer-readable storage medium. These software functional units, stored in a storage medium, include several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute partial steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes: a USB flash drive, a portable hard drive, a magnetic disk, an optical disk, volatile memory, or non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM), and direct memory bus random access memory (DRRAM), etc., which are various media that can store program code.

[0112] While the present invention has been disclosed above, it is not limited thereto. Any person skilled in the art can easily conceive of variations or substitutions without departing from the spirit and scope of the present invention, and various modifications and alterations can be made, including combinations of the different functions and implementation steps described above, as well as software and hardware implementation methods, all of which are within the protection scope of the present invention.

Claims

1. A generalized coherent quantum evidential reasoning method based on quantum evidential theory, characterized in that, include: For quantum evidence fusion scenarios containing coherent terms, multiple independent evidence sources to be combined are obtained to obtain the corresponding generalized quantum fundamental probability amplitude function, which is a quantum state characterization function that characterizes the confidence of the evidence source. The generalized quantum fundamental probability amplitude function is encoded as a quantum superposition state and initialized to a pre-constructed composite quantum system to obtain the initial quantum state; Based on the initial quantum state, and based on the generalized quantum evidence combination rule, a coherent combination quantum circuit is constructed on the composite quantum system to perform a three-stage evolution to obtain the target quantum state. The three-stage evolution sequentially performs the following operations on the initial quantum state: performing intersection operation on the focal element, erasing path information and generating a coherent superposition state, and filtering and marking valid coherent superposition terms. A two-layer quantum measurement mechanism based on post-selection projection is constructed on the composite quantum system. The generalized quantum fundamental probability measurement operator is used to perform quantum measurement on the target quantum state to extract the generalized quantum fundamental probability distribution, or the generalized quantum likelihood distribution measurement operator is used to perform quantum measurement on the target quantum state to extract the generalized quantum likelihood distribution corresponding to the single proposition subset. Data category determination is performed based on the generalized quantum fundamental probability distribution or the generalized quantum likelihood distribution.

2. The method of claim 1, wherein, For quantum evidence fusion scenarios containing coherent terms, multiple independent evidence sources to be combined are obtained to derive the corresponding generalized quantum fundamental probability amplitude function, including: For each of the evidence sources to be combined in the quantum evidence fusion scenario, obtain the normalized continuous feature data corresponding to the currently processed evidence source to be combined; Perform classical probability assignment on the normalized continuous feature data to generate the classical generalized basic probability assignment of the currently processed evidence source to be combined; The phase sign is determined based on the cardinality of the propositional subset corresponding to the classical generalized basic probability assignment, wherein the single propositional subset in the propositional subset corresponds to the positive phase sign, the multi-element composite set in the propositional subset corresponds to the negative phase sign, and the propositional subset is a subset in the identification frame; The classical generalized fundamental probability assignment is phase-adjusted according to the preset phase hyperparameter to obtain the generalized quantum fundamental probability amplitude function containing phase information of the evidence source to be combined in the current processing. The phase hyperparameter is used to control the phase angle.

3. The method of claim 2, wherein, The step of adjusting the phase of the classical generalized fundamental probability assignment according to a preset phase hyperparameter to obtain the generalized quantum fundamental probability amplitude function containing phase information of the currently processed evidence source to be combined includes: Obtain the propositional subset corresponding to the classical generalized basic probability assignment; Determine whether the cardinality of the subset of the propositions is equal to one; When the cardinality of the subset of propositions is equal to one, the phase sign is determined to be positive; When the cardinality of the subset of propositions is greater than one, the phase sign is determined to be negative; The product of the phase sign and the phase hyperparameter is used as the phase angle, which is then combined with the magnitude of the classical generalized fundamental probability assignment to generate the complex form of the generalized quantum fundamental probability magnitude function.

4. The method of claim 3, wherein, The composite quantum system includes a first evidence source data register, a second evidence source data register, a closed-world data register, an open-world indicator register, and a filter flag register. The first and second evidence source data registers are used to store the quantum-coded data of the evidence sources to be combined. The closed-world data register is used to store the closed-world focal element operation results. The open-world indicator register is used to identify open-world category information. The filter flag register is used to mark the coherent screening state. Based on the initial quantum state and the generalized quantum evidence combination rules, a coherent combination quantum circuit is constructed on the composite quantum system to perform a three-stage evolution to obtain the target quantum state, including: The focal data carried by the initial quantum state is stored in the first evidence source data register and the second evidence source data register; A computational unitary operator is applied to perform set-theoretic intersection operations on the focal data in the first evidence source data register and the second evidence source data register, and the intersection result is mapped to the closed-world data register and the open-world indicator bit register to generate an incoherent preliminary combined intermediate state. The non-empty intersection result corresponding to the intersection result is mapped to the closed-world data register, the empty intersection result is used as a conflict term, and the open-world belief term corresponds to the empty intersection result. An Adama gate operation is applied to all qubits in the first evidence source data register and the second evidence source data register to generate a superposition state containing coherent terms and auxiliary coherence indices. When the auxiliary coherence indices of the first evidence source data register and the second evidence source data register are both zero, the corresponding coherent terms are equal to the coherent orthogonal sum. The auxiliary coherence indices are binary indices used to mark the components of the superposition state. The subspace with the auxiliary coherence index of zero is flipped, and the open-world belief term and the conflict term are physically distinguished through the open-world indicator register and the closed-world data register, and the valid coherence superposition term is marked to obtain the target quantum state.

5. The method of claim 4, wherein, The process of flipping the state of the subspace where the auxiliary coherence index is zero, and physically distinguishing the open-world belief term from the conflict term using the open-world indicator bit register and the closed-world data register, and marking the valid coherent superposition term to obtain the target quantum state, includes: Obtain the superposition state containing coherent terms and auxiliary coherent indices; The state is flipped in the subspace where the auxiliary coherence index is zero by a filtering operator; The state of the open-world indicator bit register is identified in the superposition component after the state flip; The state of the closed-world data register is identified in the superposition component after the state flip; Based on the state of the open-world indicator register and the state of the closed-world data register, the open-world belief term and the conflict term are distinguished as different quantum state components; The target quantum state is obtained by marking the constructive coherence terms containing the correct fusion information using the filtering operator.

6. The method of claim 4, wherein, The construction of a two-layer quantum measurement mechanism based on post-selection projection on the composite quantum system, which extracts the generalized quantum fundamental probability distribution by performing quantum measurement on the target quantum state using a generalized quantum fundamental probability measurement operator, or extracts the generalized quantum likelihood distribution corresponding to a single proposition subset by performing quantum measurement on the target quantum state using a generalized quantum likelihood distribution measurement operator, includes: After passing through, the projection operator is selected to perform projection measurement on the filter flag register to remove noise samples that do not meet the zero-frequency coherence condition, and obtain an effective sampling ensemble; On the effective sampling ensemble, either a confidence layer measurement or a likelihood layer measurement is selected to be performed; if the confidence layer measurement is performed, the generalized quantum fundamental probability measurement operator is used to perform a joint measurement on the closed-world data register and the open-world indicator bit register to obtain the sampling count corresponding to the proposition subset, which is used as the generalized quantum fundamental probability distribution; If the likelihood layer measurement is performed, the generalized quantum likelihood distribution measurement operator is used to perform an edge measurement on a single bit of the closed-world data register to obtain the edge excitation count corresponding to the single proposition subset, which is used as the generalized quantum likelihood distribution corresponding to the single proposition subset.

7. The method of claim 6, wherein, The process of determining the data category based on the generalized quantum fundamental probability distribution or the generalized quantum likelihood distribution includes: The subset of propositions corresponding to the maximum value in the sampled count is determined as the classification result based on the generalized quantum fundamental probability distribution; or... The subset of single propositions corresponding to the maximum value in the edge excitation count is determined as the classification result based on the generalized quantum likelihood distribution. When the maximum value corresponds to an empty set, an open-world warning is triggered, and the set is identified as an unknown new category.

8. A generalized coherent quantum evidential reasoning device based on quantum evidential theory, characterized in that, include: The function determination unit is used to obtain multiple independent evidence sources to be combined for a quantum evidence fusion scenario containing coherent terms, so as to obtain the corresponding generalized quantum fundamental probability amplitude function, wherein the generalized quantum fundamental probability amplitude function is a quantum state characterization function that characterizes the confidence of the evidence source. An initialization unit is used to encode the generalized quantum fundamental probability amplitude function into a quantum superposition state and initialize it to a pre-constructed composite quantum system to obtain an initial quantum state; The combinatorial evolution unit is used to construct a coherent combinatorial quantum circuit on the composite quantum system based on the initial quantum state and the generalized quantum evidence combinatorial rule, so as to perform a three-stage evolution to obtain the target quantum state. The three-stage evolution sequentially performs the following operations on the initial quantum state: performing an intersection operation on the focal element, erasing path information and generating a coherent superposition state, and filtering and marking valid coherent superposition terms. A quantum measurement unit is used to construct a two-layer quantum measurement mechanism based on post-selection projection on the composite quantum system, and to perform quantum measurement on the target quantum state using a generalized quantum fundamental probability measurement operator to extract the generalized quantum fundamental probability distribution, or to perform quantum measurement on the target quantum state using a generalized quantum likelihood distribution measurement operator to extract the generalized quantum likelihood distribution corresponding to a single proposition subset. The classification decision unit is used to perform data category determination based on the generalized quantum fundamental probability distribution or the generalized quantum likelihood distribution.

9. An electronic device, characterized in that, The method includes a processor, a memory, a communication interface, and one or more programs, said programs being stored in the memory and configured to be executed by the processor, said programs including instructions for performing the steps of the method as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, A computer program for storing electronic data interchange is provided, wherein the computer program causes a computer to perform the method as described in any one of claims 1-7.

Images