Train speed control method and device based on data encryption, equipment and medium

Abstract

The invention discloses a train speed control method, device and equipment based on data encryption and a medium, and the method comprises the steps: calculating the deviation between the actual running speed and the expected speed of a target train, and obtaining a current speed tracking error; if the speed tracking error is greater than a preset trigger threshold value, calculating a first control update item according to the control gain and the speed tracking error which are calculated on line, and encrypting the first control update item to obtain a second control update item; the second control update item is uploaded to a cloud end, so that the control input ciphertext of the current iteration is obtained from the cloud end, and the control input plaintext of the current iteration is obtained through decryption, so that the train speed of the target train is controlled; wherein the control input ciphertext of the current iteration is obtained by updating the control input ciphertext of the previous iteration by the cloud end by using homomorphic addition characteristics. According to the method, the real-time performance of train speed control can be improved while the data security is ensured through local and cloud hybrid data encryption.

Description

Technical Field

[0001] This invention relates to the field of train speed control technology, and in particular to a train speed control method, device, equipment and medium based on data encryption. Background Technology

[0002] As automated train operation systems (AEBS) evolve towards greater autonomy and networking, critical data generated during train speed control requires frequent interaction between onboard terminals and cloud-based dispatch centers via public networks. This data is central to the safe and efficient operation of trains and constitutes sensitive information within critical infrastructure. Theft or tampering during transmission and processing would directly jeopardize both train operation and public safety. Therefore, encrypting and protecting the data involved in train speed control processes to prevent leakage in untrusted network environments and computational stages has become an indispensable technical requirement for ensuring the cybersecurity of modern rail transit systems.

[0003] Currently, existing technologies typically employ homomorphic encryption to protect control data. Specifically, a typical implementation involves executing the critical computational steps of the iterative learning control law entirely in the cloud. Without decryption, the cloud directly performs homomorphic multiplication on the received encrypted control gain and encrypted tracking error, thus completing the iterative update of the control input in encrypted form. However, this approach has a significant technical drawback: homomorphic multiplication is extremely computationally complex, often taking several orders of magnitude longer than homomorphic addition. When this computationally intensive homomorphic multiplication is deployed in a train speed control closed loop with extremely stringent real-time requirements, it introduces unacceptably large computational delays. This delay not only reduces the response speed of the control system and affects tracking accuracy but may also lead to security risks in extreme cases due to untimely updates of control commands. Therefore, how to effectively encrypt and protect train speed control data while overcoming the real-time bottleneck caused by encrypted computation (especially homomorphic multiplication) has become a core technical problem urgently needing to be solved in this field. Summary of the Invention

[0004] This invention provides a train speed control method, device, equipment, and medium based on data encryption, which can improve the real-time performance of train speed control while ensuring data security through hybrid local and cloud data encryption.

[0005] In a first aspect, embodiments of the present invention provide a train speed control method based on data encryption, applicable to train speed iterative learning control systems, comprising: The actual operating speed of the target train at the current moment is collected, and the deviation between the actual operating speed and the preset expected speed is calculated to obtain the speed tracking error at the current moment; If the speed tracking error is greater than a preset trigger threshold, then a first control update term is calculated based on the online calculated control gain and the speed tracking error, and the first control update term is encrypted using a preset Paillier public key encryption algorithm to obtain a second control update term; The second control update item is uploaded to the cloud to obtain the control input ciphertext from the cloud, and the control input ciphertext is decrypted to obtain the control input plaintext of the target train in the current iteration, so as to control the train speed of the target train through the control input plaintext of the current iteration; wherein, the control input ciphertext of the current iteration is obtained by updating the control input ciphertext of the previous iteration in the cloud using the homomorphic addition property.

[0006] This invention ensures the accuracy of speed control data during iterative learning and control of train speed by calculating speed tracking errors. A threshold trigger mechanism ensures that the system only initiates subsequent high-overhead encrypted communication processes when control performance is poor (large errors), avoiding unnecessary computational overhead. Furthermore, by performing complex multiplication operations on plaintext locally beforehand, the update items are updated, significantly reducing computational latency compared to traditional methods that perform homomorphic multiplication on encrypted text in the cloud. By directly performing homomorphic addition (which is far less computationally complex than homomorphic multiplication) on the encrypted update items in the cloud, high-overhead multiplication calculations are prevented from occurring in the cloud, further guaranteeing the real-time nature of control input acquisition. Compared to existing technologies, this invention improves the real-time performance of train speed control while ensuring data security through hybrid local and cloud data encryption.

[0007] In some preferred embodiments of the first aspect, the step of calculating a first control update term based on the online calculated control gain and the speed tracking error, and encrypting the first control update term using a preset Paillier public-key encryption algorithm to obtain a second control update term, specifically involves: Based on the plaintext of the control input and the plaintext of the control output from the two most recent historical iterations, calculate the control gain, and based on the control gain and the speed tracking error, calculate the first control update term; The first control update item is encrypted using a preset Paillier public key encryption algorithm to obtain the second control update item.

[0008] This invention strengthens privacy protection at the data level by encrypting the two core sensitive data, control gain and velocity tracking error, in plaintext locally to obtain a second control update item. This makes the subsequent encryption object a single update item that has integrated sensitive information (error and implicit dynamic gain), rather than the original, separate sensitive data.

[0009] In some preferred embodiments of the first aspect, the control gain is calculated based on the plaintext of the control input and the plaintext of the control output from the two most recent historical iterations, specifically as follows: Based on the plaintext of the control input in the historical iterations, calculate the change in control input between the two most recent historical iterations; Based on the control output plaintext of the historical iterations, calculate the velocity change between the two most recent historical iterations; Based on the change in control input and the change in velocity, the pseudo-partial derivatives are updated using a preset recursive least squares algorithm, and the control gain is calculated based on the updated pseudo-partial derivatives; wherein, the estimated pseudo-partial derivatives are used to characterize the dynamic characteristics of the system.

[0010] The embodiments of the present invention calculate the control gain through a preset recursive least squares algorithm. By adopting the principle of data-driven, model-free adaptation, it can learn online and adapt to the dynamic changes of the train without relying on a fixed physical model, thus ensuring the high-precision tracking performance of the control system under various disturbances.

[0011] In some preferred embodiments of the first aspect, the step of encrypting the first control update item using a preset Paillier public-key encryption algorithm to obtain the second control update item specifically involves: The first control update term is quantized and mapped to an integer to obtain the first control update term in integer form. The first control update item in integer form is encrypted using the public key of a preset homomorphic encryption system to generate the second control update item.

[0012] This invention converts update terms of the control field (real number) into processable data of the cryptographic field (integer).

[0013] In some preferred embodiments of the first aspect, the control input ciphertext of the current iteration is obtained by updating the pre-stored control input ciphertext of the previous iteration in the cloud using the homomorphic addition property, based on the second control update term.

[0014] This invention eliminates the bottleneck of homomorphic multiplication in the cloud by performing homomorphic addition with minimal computational overhead, ensuring the overall real-time performance of the system. Furthermore, the cloud cannot know any plaintext information during this process, further improving the security of control command iteration updates.

[0015] In some preferred embodiments of the first aspect, the control input ciphertext is decrypted to obtain the control input plaintext for the current iteration of the target train, specifically as follows: The control input ciphertext is decrypted using the private key corresponding to the public key in the homomorphic encryption system to obtain the control input in integer form; The integer form of the control input is subjected to inverse mapping and inverse quantization to obtain the real number form of the control input plaintext.

[0016] The embodiments of the present invention ensure the feasibility of decryption and control by converting from the cryptographic domain to the control domain.

[0017] Among the preferred options in the first aspect are: If the speed tracking error is less than or equal to the trigger threshold, the control input plaintext of the previous iteration is directly used as the control input plaintext of the current iteration to control the train speed of the target train.

[0018] This invention, by completely skipping all complex steps such as calculation, encryption, and cloud interaction when not triggered, greatly saves local computing resources and network bandwidth, avoids any unnecessary data exposure during this period, and maintains the continuity of control.

[0019] Secondly, embodiments of the present invention provide a train speed control device based on data encryption, suitable for train speed iterative learning control systems, including a speed error acquisition module, an update term acquisition module, and a control input acquisition module, wherein... The speed error acquisition module is used to collect the actual running speed of the target train at the current moment, and calculate the deviation between the actual running speed and the preset expected speed to obtain the speed tracking error at the current moment. The update item acquisition module is used to calculate a first control update item based on the online calculated control gain and the speed tracking error if the speed tracking error is greater than a preset trigger threshold, and to encrypt the first control update item using a preset Paillier public key encryption algorithm to obtain a second control update item. The control input acquisition module is used to upload the second control update item to the cloud, obtain the control input ciphertext from the cloud, and decrypt the control input ciphertext to obtain the control input plaintext of the target train in the current iteration, so as to control the train speed of the target train through the control input plaintext of the current iteration; wherein, the control input ciphertext of the current iteration is obtained by updating the control input ciphertext of the previous iteration in the cloud using the homomorphic addition property.

[0020] This invention employs a speed error acquisition module to calculate speed tracking error, ensuring the accuracy of speed control data during train speed iterative learning control. An update item acquisition module introduces a threshold trigger mechanism, ensuring that the system only initiates subsequent high-overhead encrypted communication processes when control performance is poor (large errors), avoiding unnecessary computational overhead. Furthermore, by performing complex multiplication operations on plaintext locally beforehand on the update item, it significantly reduces computational latency compared to traditional methods that perform homomorphic multiplication on ciphertext in the cloud. Finally, a control input acquisition module acquires the control input ciphertext encrypted by performing homomorphic addition (a computationally much less complex operation than homomorphic multiplication) on the encrypted update item in the cloud, further improving computational efficiency and ensuring the real-time acquisition of control input for the current iteration.

[0021] Thirdly, embodiments of the present invention provide a terminal device, including: a processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other through the communication bus; The memory is used to store at least one executable instruction that causes the processor to perform the operation of the data encryption-based train speed control method as described in any of the preceding descriptions.

[0022] Fourthly, embodiments of the present invention provide a computer-readable storage medium, the computer-readable storage medium including a stored computer program, wherein, when the computer program is executed, it controls the device or apparatus where the computer-readable storage medium is located to perform the data encryption-based train speed control method as described in any of the preceding claims.

[0023] The above description is merely an overview of the technical solutions of the embodiments of the present invention. In order to better understand the technical means of the embodiments of the present invention and to implement them in accordance with the contents of the specification, and to make the above and other objects, features and advantages of the embodiments of the present invention more apparent and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0024] Figure 1 A schematic diagram of a train speed control method based on data encryption provided in an embodiment of the present invention; Figure 2 A flowchart illustrating a train speed control based on data encryption, as exemplified by an embodiment of the present invention. Figure 3 This is a structural diagram of a train speed control device based on data encryption, provided for an embodiment of the present invention. Detailed Implementation

[0025] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0026] Example 1: like Figure 1 As shown, an embodiment of the present invention provides a train speed control method based on data encryption, applicable to train speed iterative learning control systems, comprising the following steps: S101, Collect the actual running speed of the target train at the current moment, and calculate the deviation between the actual running speed and the preset expected speed to obtain the speed tracking error at the current moment; It should be noted that the train speed iterative learning control system utilizes ILC (Iterative Learning Control Algorithm) to simulate the human "learning" process. Based on the tracking information from the previous iteration, it continuously corrects the control input signal for the current iteration, achieving complete trajectory tracking over a fixed interval. Compared with traditional control methods, data-driven ILC has the following two core advantages: 1) It does not require knowledge of the system's precise mathematical model, enabling its application to train dynamic systems with strong nonlinearity and uncertainty; 2) It can achieve high-precision tracking throughout the entire operating range.

[0027] In one specific embodiment, the iterative learning control algorithm is implemented as follows: The dynamic mathematical model of a class of high-speed trains moving on a straight track can be represented as: in (s) represents the train travel time. (m / s) represents the operating speed of the HST, and the control input F(t) (N / kg) represents the traction force in terms of mass. (N / kg) is the basic resistance per unit mass, and s(t)(m) is the operating position of the HST (High-Speed ​​Train). It is the basic drag coefficient. It is the additional drag coefficient. Indicates position The equivalent tilt angle at that point, This indicates additional resistance.

[0028] Additional resistance This can be expressed as: Furthermore, the above dynamic mathematical model can be discretized using Euler methods to obtain the discretized train dynamics model: Where, j This represents a discrete-time index. (s) represents the sampling time, T= J. = , .

[0029] Based on the discretized train dynamics model, the control input... It can be divided into the following two parts: = in, = This is for the last two terms in the discretized train dynamics model, and This is the part that needs to be designed. Further, the decomposed control input... Will be rewritten as By speed and control Defined as the system's output and input respectively, the previous formula (decomposed control input) This can be rewritten as: in, and These are the system output and input at time j during the k-th iteration, respectively. = It is an unknown but bounded parameter vector. =[ ], >0 controls the input gain.

[0030] Therefore, the train model can be equivalent to a general repeatable discrete-time nonlinear system, and the nonlinear function f( (Unknown) The discrete nonlinear dynamic system is as shown in the following formula: For the discrete nonlinear system (i.e., the third formula, the discretized train dynamics model), the following two assumptions are proposed: Assumption 1: Nonlinear function f Input to train It is continuously differentiable.

[0031] Assumption 2: For ∀j∈ and k∈ System (3) satisfies the generalized Lipsitz condition, i.e. for >0 is a constant. This represents the difference between two consecutive iterations.

[0032] Assumption 3: Total quantization error caused by encrypted links It is bounded.

[0033] In this embodiment, the method further includes: if the speed tracking error is less than or equal to the trigger threshold, then the control input plaintext of the previous iteration is directly used as the control input plaintext of the current iteration to control the train speed of the target train.

[0034] In one possible embodiment, the desired speed is obtained through offline planning and pre-storage: Before train operation, an optimal speed-position curve or speed-time curve from the starting point to the destination is calculated using algorithms, based on constraints such as track conditions (gradient, curvature, speed limits), timetable (inter-station time, energy-saving requirements), and passenger comfort. This optimized curve is pre-stored in the onboard controller or a connected safety computer. During operation, the system only needs to query or interpolate the desired speed value at the corresponding moment from the stored curve based on the current train position or time.

[0035] In one possible embodiment, the desired speed is received in real time from the dispatch center: In a more advanced networked operation mode, the ground dispatch center can dynamically generate or fine-tune the desired speed curve based on real-time road conditions, the positions of other trains, temporary speed limit instructions, etc., and send it to the train through a secure data communication network. After receiving and verifying the instruction, the train's onboard system uses it as the current control target.

[0036] S102, if the speed tracking error is greater than a preset trigger threshold, then calculate a first control update item based on the online calculated control gain and the speed tracking error, and encrypt the first control update item using a preset Paillier public key encryption algorithm to obtain a second control update item; In this embodiment, the step of calculating a first control update term based on the online calculated control gain and the speed tracking error, and encrypting the first control update term using a preset Paillier public key encryption algorithm to obtain a second control update term, specifically involves: calculating the control gain based on the plaintext of the control input and the plaintext of the control output from the two most recent historical iterations; calculating the first control update term based on the control gain and the speed tracking error; and encrypting the first control update term using a preset Paillier public key encryption algorithm to obtain the second control update term.

[0037] In one specific embodiment, for the discrete nonlinear dynamic system, given that To address the unknowns, we propose the following data-driven adaptive iterative learning control method, whose update law incorporates a threshold triggering mechanism: in, The train sensor monitors the actual running speed at the current time j in real time, based on a preset trigger threshold, and outputs... And the total applied traction force, which is the input. The local controller calculates the deviation from the preset target speed. Using the plaintext input and output data generated in the previous two iterations, the pseudo-partial derivatives are updated locally in real time using the recursive least squares method. This leads to the adaptive control gain. , This is the total quantitative error of the system. If... This indicates that the train speed deviates too far from the target curve (e.g., encountering sudden high wind resistance or a change in gradient). In this case, a learning update must be activated, entering the cloud-based interactive process to quickly correct the control parameters. If This indicates that the current control performance is good. The local controller uses the control input from the previous iteration. This interrupts communication and computing with the cloud, saving communication and computing resources and ensuring the stability and efficiency of train operation.

[0038] In this embodiment, the control gain is calculated based on the plaintext control input and control output of the two most recent historical iterations. Specifically, this involves: calculating the change in control input between the two most recent historical iterations based on the plaintext control input of the historical iterations; calculating the change in velocity between the two most recent historical iterations based on the plaintext control output of the historical iterations; updating the pseudo-partial derivatives using a preset recursive least squares algorithm based on the change in control input and the change in velocity; and calculating the control gain based on the updated pseudo-partial derivatives. The estimated pseudo-partial derivatives are used to characterize the dynamic characteristics of the system.

[0039] In one specific embodiment, gain The calculation does not depend on Instead of providing a precise expression, it iteratively updates the data by observing the input and output data of the observation system, for example, through the following pseudo-partial derivative estimation algorithm. Then calculate , in = - , = - , , , , These are the preset positive learning parameters.

[0040] For the discrete nonlinear dynamic system, under the conditions of satisfying assumptions 1, 2, and 3, and , 0,0 In the case of applying data-driven adaptive iterative learning control law (12), if the adaptive control gain It converges online to an appropriate interval and satisfies: The tracking error of the discrete nonlinear dynamic system is then addressed. During the iteration process, it will eventually converge to a point where Within the neighborhood bounded by the quantization error.

[0041] In this embodiment, the first control update item is encrypted using a preset Paillier public key encryption algorithm to obtain the second control update item. Specifically, the first control update item is quantized and mapped to an integer to obtain an integer form of the first control update item; the integer form of the first control update item is encrypted using the public key of a preset homomorphic encryption system to generate the second control update item.

[0042] S103, the second control update item is uploaded to the cloud to obtain the control input ciphertext from the cloud, and the control input ciphertext is decrypted to obtain the control input plaintext of the target train in the current iteration, so as to control the train speed of the target train through the control input plaintext of the current iteration; wherein, the control input ciphertext of the current iteration is obtained by updating the control input ciphertext of the previous iteration by the cloud using the homomorphic addition property.

[0043] In one specific embodiment, the encryption process is as follows: By randomly selecting two that satisfy gcd( Two large prime numbers whose sum is 1 and Public keys can be established. and private key A homomorphic cryptosystem where the public key is used to encrypt system signals and the private key is used to decrypt signals. For any plaintext... Encryption operations can be given as Where r is from the set .

[0044] In this embodiment, the control input ciphertext of the current iteration is obtained by updating the pre-stored control input ciphertext of the previous iteration in the cloud using the homomorphic addition property according to the second control update item.

[0045] In this embodiment, the control input ciphertext is decrypted to obtain the control input plaintext for the current iteration of the target train. Specifically, the control input ciphertext is decrypted using the private key corresponding to the public key in the homomorphic encryption system to obtain the control input in integer form; the integer form control input is then subjected to inverse mapping and inverse quantization to obtain the control input plaintext in real number form.

[0046] In one specific embodiment, for ciphertext c The corresponding decryption operation is defined as follows: in, yes The inverse of the modulus.

[0047] Preferred, definition = The objective function is given below: Used to balance “tracking error” and “control input variation”.

[0048] In one specific embodiment, when the threshold is triggered, i.e. This indicates that the train speed deviates too far from the target curve (e.g., encountering sudden high wind resistance or a change in gradient). At this point, a learning update must be activated, entering the cloud-based interactive process to quickly correct the control inputs. The onboard controller obtains adaptive control gain by collecting real-time speed and traction data. And this gain is compared with the velocity tracking error. Multiply to obtain the updated term, then quantize, map, and encrypt the updated term on the local trusted controller. Upload the encrypted updated term to the cloud controller, and then perform homomorphic addition with the previously obtained input ciphertext to obtain the new input. in, It is the ciphertext of the control input at time j in the k-th iteration. It is the ciphertext of the update term obtained by multiplying the adaptive gain and error at time j of the k-th iteration.

[0049] Preferably, in the above recursive update formula, the parameter and These parameters act as the learning rate and weighting factor. The adjustment and calculation of these parameters are performed entirely locally on the onboard controller in plaintext to enable rapid adaptation to the time-varying characteristics of train dynamics.

[0050] Preferably, to prevent spurious partial derivatives During the learning process, the estimated values ​​may diverge or exhibit abnormal deviations due to train coupling or emergency braking. To address this, we introduce a parameter reset mechanism. This mechanism is executed entirely in plaintext on a local trusted terminal, without involving cloud-based interactions with sensitive data, thus ensuring the autonomous and safe operation of the onboard system.

[0051] Preferably, the termination condition of the adaptive learning mechanism is consistent with the termination condition of the entire iterative learning control loop. The adaptive learning process ends when the number of iterations of the iterative learning algorithm reaches the preset maximum number of iterations, `max_iter` (e.g., after completing 100 runs on a specific route). At this point, the latest adaptive control gain... and its corresponding pseudo-partial derivatives This is considered to be the optimization result within the current iteration cycle, the result of online optimization for the dynamic characteristics of the line, used to generate the high-precision control law that eventually converges.

[0052] For a better explanation of the working principle and process of this invention, please refer to [link / reference]. Figure 2 , Figure 2 This is an example of a train speed control flowchart based on data encryption, as exemplified by an embodiment of the present invention.

[0053] This invention ensures the accuracy of speed control data during iterative learning control of train speed by calculating speed tracking errors. A threshold trigger mechanism ensures that the system only initiates subsequent high-overhead encrypted communication processes when control performance is poor (large errors), avoiding unnecessary computational overhead. Furthermore, by performing complex multiplication operations on plaintext locally beforehand on the update items, it significantly reduces computational latency compared to traditional methods that perform homomorphic multiplication on ciphertext in the cloud. By directly performing homomorphic addition on the encrypted update items in the cloud (homomorphic addition is far less computationally complex than homomorphic multiplication), computational efficiency is further improved, ensuring the real-time acquisition of the current iterative control input. Compared to existing technologies, this invention improves the real-time performance of train speed control while ensuring data security through hybrid local and cloud data encryption.

[0054] Example 2: like Figure 3 As shown, this embodiment provides a train speed control device based on data encryption, suitable for train speed iterative learning control systems. It includes a speed error acquisition module 201, an update item acquisition module 202, and a control input acquisition module 203. The speed error acquisition module 201 is used to collect the actual running speed of the target train at the current moment, and calculate the deviation between the actual running speed and the preset expected speed to obtain the speed tracking error at the current moment. The update item acquisition module 202 is used to calculate a first control update item based on the online calculated control gain and the speed tracking error if the speed tracking error is greater than a preset trigger threshold, and to encrypt the first control update item using a preset Paillier public key encryption algorithm to obtain a second control update item; wherein, the train speed control plaintext includes control input plaintext and control output plaintext; In this embodiment, the update item acquisition module 202 calculates a first control update item based on the online calculated control gain and the speed tracking error, and encrypts the first control update item using a preset Paillier public key encryption algorithm to obtain a second control update item. Specifically, the update item acquisition module 202 calculates the control gain based on the control input plaintext and control output plaintext of the two most recent historical iterations, and calculates the first control update item based on the control gain and the speed tracking error; the first control update item is then encrypted using a preset Paillier public key encryption algorithm to obtain the second control update item.

[0055] In this embodiment, the update item acquisition module 202 calculates the control gain based on the plaintext control input and control output of the two most recent historical iterations. Specifically, the update item acquisition module 202 calculates the change in control input between the two most recent historical iterations based on the plaintext control input of the historical iterations; calculates the change in velocity between the two most recent historical iterations based on the plaintext control output of the historical iterations; updates the pseudo-partial derivatives using a preset recursive least squares algorithm based on the change in control input and the change in velocity, and calculates the control gain based on the updated pseudo-partial derivatives; wherein the estimated value of the pseudo-partial derivatives is used to characterize the dynamic characteristics of the system.

[0056] In this embodiment, the update item acquisition module 202 encrypts the first control update item using a preset Paillier public key encryption algorithm to obtain the second control update item. Specifically, the update item acquisition module 202 performs quantization and integer mapping processing on the first control update item to obtain the first control update item in integer form; and uses the public key of a preset homomorphic encryption system to encrypt the first control update item in integer form to generate the second control update item.

[0057] The control input acquisition module 203 is used to upload the second control update item to the cloud, obtain the control input ciphertext from the cloud, and decrypt the control input ciphertext to obtain the control input plaintext of the target train in the current iteration, so as to control the train speed of the target train through the control input plaintext of the current iteration; wherein, the control input ciphertext of the current iteration is obtained by updating the control input ciphertext of the previous iteration by the cloud using the homomorphic addition property.

[0058] In this embodiment, the control input acquisition module 203 decrypts the ciphertext of the control input to obtain the plaintext of the control input for the current iteration of the target train. Specifically, the control input acquisition module 203 decrypts the ciphertext of the control input using the private key corresponding to the public key in the homomorphic encryption system to obtain the control input in integer form; and performs inverse mapping and inverse quantization processing on the integer form of the control input to obtain the plaintext of the control input in real number form.

[0059] For a more detailed explanation of the working principle and procedures of this embodiment, please refer to the relevant description in Embodiment 1.

[0060] This invention employs a speed error acquisition module 201 to calculate speed tracking error, ensuring the accuracy of speed control data during the train speed iterative learning control process. An update item acquisition module 202 introduces a threshold trigger mechanism, ensuring that the system only initiates subsequent high-overhead encrypted communication processes when control performance is poor (large errors), avoiding unnecessary computational overhead. Furthermore, by performing complex multiplication operations on the plaintext locally beforehand, the computational latency is significantly reduced compared to the traditional method of performing homomorphic multiplication on ciphertext in the cloud. Finally, a control input acquisition module 203 acquires the control input ciphertext encrypted by performing homomorphic addition (which is far less computationally complex than homomorphic multiplication) on the encrypted update item in the cloud, further improving computational efficiency and ensuring the real-time acquisition of the current iteration's control input.

[0061] Example 3: This embodiment provides a terminal device, including: a processor, a memory, a communication interface, and a communication bus. The processor, the memory, and the communication interface communicate with each other through the communication bus. The memory is used to store at least one executable instruction, which causes the processor to perform the operation of the data encryption-based train speed control method as described above.

[0062] Example 4: This invention provides a computer-readable storage medium comprising a stored computer program, wherein, when the computer program is executed, it controls the device or apparatus containing the computer-readable storage medium to perform the data encryption-based train speed control method as described above.

[0063] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0064] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above descriptions are merely specific embodiments of the present invention and are not intended to limit the scope of protection of the present invention. In particular, it should be noted that any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention for those skilled in the art.

Claims

1. A train speed control method based on data encryption, characterized in that, Applicable to train speed iterative learning control systems, including: The actual operating speed of the target train at the current moment is collected, and the deviation between the actual operating speed and the preset expected speed is calculated to obtain the speed tracking error at the current moment; If the speed tracking error is greater than a preset trigger threshold, then a first control update term is calculated based on the online calculated control gain and the speed tracking error, and the first control update term is encrypted using a preset Paillier public key encryption algorithm to obtain a second control update term; The second control update item is uploaded to the cloud to obtain the control input ciphertext from the cloud, and the control input ciphertext is decrypted to obtain the control input plaintext of the target train in the current iteration, so as to control the train speed of the target train through the control input plaintext of the current iteration; wherein, the control input ciphertext of the current iteration is obtained by updating the control input ciphertext of the previous iteration by the cloud using the homomorphic addition property.

2. The train speed control method based on data encryption as described in claim 1, characterized in that, The first control update term is calculated based on the online calculated control gain and the speed tracking error. This first control update term is then encrypted using a preset Paillier public-key encryption algorithm to obtain the second control update term. Specifically: Based on the plaintext of the control input and the plaintext of the control output from the two most recent historical iterations, calculate the control gain, and based on the control gain and the speed tracking error, calculate the first control update term; The first control update item is encrypted using a preset Paillier public key encryption algorithm to obtain the second control update item.

3. The train speed control method based on data encryption as described in claim 2, characterized in that, Based on the plaintext of the control input and control output from the two most recent historical iterations, the control gain is calculated as follows: Based on the plaintext of the control input in the historical iterations, calculate the change in control input between the two most recent historical iterations; Based on the control output plaintext of the historical iterations, calculate the velocity change between the two most recent historical iterations; Based on the change in control input and the change in velocity, the pseudo-partial derivatives are updated using a preset recursive least squares algorithm, and the control gain is calculated based on the updated pseudo-partial derivatives; wherein, the estimated pseudo-partial derivatives are used to characterize the dynamic characteristics of the system.

4. The train speed control method based on data encryption as described in claim 2, characterized in that, The first control update item is encrypted using a preset Paillier public key encryption algorithm to obtain the second control update item, specifically as follows: The first control update term is quantized and mapped to an integer to obtain the first control update term in integer form. The first control update item in integer form is encrypted using the public key of a preset homomorphic encryption system to generate the second control update item.

5. The train speed control method based on data encryption as described in claim 1, characterized in that, The control input ciphertext for the current iteration is obtained by updating the pre-stored control input ciphertext of the previous iteration in the cloud using the homomorphic addition property, based on the second control update term.

6. The train speed control method based on data encryption as described in claim 4, characterized in that, Decrypting the ciphertext of the control input yields the plaintext of the control input for the current iteration of the target train, specifically: The control input ciphertext is decrypted using the private key corresponding to the public key in the homomorphic encryption system to obtain the control input in integer form; The integer form of the control input is subjected to inverse mapping and inverse quantization to obtain the real number form of the control input plaintext.

7. The train speed control method based on data encryption as described in claim 1, characterized in that, Also includes: If the speed tracking error is less than or equal to the trigger threshold, the control input plaintext of the previous iteration is directly used as the control input plaintext of the current iteration to control the train speed of the target train.

8. A train speed control device based on data encryption, characterized in that, This is applicable to train speed iterative learning control systems, including a speed error acquisition module, an update term acquisition module, and a control input acquisition module. The speed error acquisition module is used to collect the actual running speed of the target train at the current moment, and calculate the deviation between the actual running speed and the preset expected speed to obtain the speed tracking error at the current moment. The update item acquisition module is used to calculate a first control update item based on the online calculated control gain and the speed tracking error if the speed tracking error is greater than a preset trigger threshold, and to encrypt the first control update item using a preset Paillier public key encryption algorithm to obtain a second control update item. The control input acquisition module is used to upload the second control update item to the cloud, obtain the control input ciphertext from the cloud, and decrypt the control input ciphertext to obtain the control input plaintext of the target train in the current iteration, so as to control the train speed of the target train through the control input plaintext of the current iteration; wherein, the control input ciphertext of the current iteration is obtained by updating the control input ciphertext of the previous iteration in the cloud using the homomorphic addition property.

9. A terminal device, characterized in that, include: The processor, memory, communication interface, and communication bus are provided, wherein the processor, memory, and communication interface communicate with each other via the communication bus. The memory is used to store at least one executable instruction that causes the processor to perform the operation of the data encryption-based train speed control method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed, it controls the device or apparatus containing the computer-readable storage medium to perform the data encryption-based train speed control method as described in any one of claims 1 to 7.

Images